3W dataset's General Presentation

This is a general presentation of the 3W dataset, to the best of its authors' knowledge, the first realistic and public dataset with rare undesirable real events in oil wells that can be readily used as a benchmark dataset for development of machine learning techniques related to inherent difficulties of actual data.

For more information about the theory behind this dataset, refer to the paper A Realistic and Public Dataset with Rare Undesirable Real Events in Oil Wells published in the Journal of Petroleum Science and Engineering (link here).

1. Introduction

This Jupyter Notebook presents the 3W dataset in a general way. For this, some tables, graphs, and statistics are presented.

2. Imports and Configurations

import sys
import os
import pathlib
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import re

import tensorflow as tf
#NUM_PARALLEL_EXEC_UNITS = 6
#config = tf.compat.v1.ConfigProto(intra_op_parallelism_threads = NUM_PARALLEL_EXEC_UNITS, 
#         inter_op_parallelism_threads = 1, 
#         allow_soft_placement = True, 
#         device_count = {'CPU': NUM_PARALLEL_EXEC_UNITS })
#session = tf.compat.v1.Session(config=config)
#tf.compat.v1.keras.backend.set_session(session)

from tensorflow import keras
from tensorflow.keras import layers
from tensorflow.keras import regularizers
import tensorflow_addons as tfa

from util_3WT import d3w, CustomDataGen

import pickle
import bisect
import sklearn
import sklearn.model_selection
from sklearn.metrics import classification_report, ConfusionMatrixDisplay, multilabel_confusion_matrix, confusion_matrix
from sklearn.utils.class_weight import compute_class_weight
%matplotlib inline
%config InlineBackend.figure_format = 'svg'

from tensorflow.python.client import device_lib
print(device_lib.list_local_devices())

[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 4132666557891900372
xla_global_id: -1
, name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 1422540800
locality {
  bus_id: 1
  links {
  }
}
incarnation: 13611297491031944386
physical_device_desc: "device: 0, name: NVIDIA GeForce GTX 950, pci bus id: 0000:01:00.0, compute capability: 5.2"
xla_global_id: 416903419
]

3. Instances' Structure

Below, all 3W dataset's instances are loaded and the first one of each knowledge source (real, simulated and hand-drawn) is partially displayed.

if pathlib.Path('dset.pkl').exists():
  with open('dset.pkl', 'rb') as f:
    dset = pickle.load(f)
else:
    dset = d3w('../dataset')
    with open('dset.pkl', 'wb') as f:
      pickle.dump(dset, f)

flist0 = ['P-PDG', 'P-TPT', 'T-TPT', 'P-MON-CKP', 'T-JUS-CKP', 'P-JUS-CKGL', 'T-JUS-CKGL', 'QGL']
categories=[[0,1,2,3,4,5,6,7,8,101,102,103,104,105,106,107,108]]

train_df, test_df, val_df = dset.split(real=True, simul=True, drawn=True, test_size=0.1, val_size=0.2, sample_n=300)

Instances Train: 213  Test: 30  Validation: 54

Each instance is stored in a CSV file and loaded into a pandas DataFrame. Each observation is stored in a line in the CSV file and loaded as a line in the pandas DataFrame. The first line of each CSV file contains a header with column identifiers. Each column of CSV files stores the following type of information:

• timestamp: observations timestamps loaded into pandas DataFrame as its index;
• P-PDG: pressure variable at the Permanent Downhole Gauge (PDG);
• P-TPT: pressure variable at the Temperature and Pressure Transducer (TPT);
• T-TPT: temperature variable at the Temperature and Pressure Transducer (TPT);
• P-MON-CKP: pressure variable upstream of the production choke (CKP);
• T-JUS-CKP: temperature variable downstream of the production choke (CKP);
• P-JUS-CKGL: pressure variable upstream of the gas lift choke (CKGL);
• T-JUS-CKGL: temperature variable upstream of the gas lift choke (CKGL);
• QGL: gas lift flow rate;
• class: observations labels associated with three types of periods (normal, fault transient, and faulty steady state).

Other information are also loaded into each pandas Dataframe:

• label: instance label (event type);
• well: well name. Hand-drawn and simulated instances have fixed names. Real instances have names masked with incremental id;
• id: instance identifier. Hand-drawn and simulated instances have incremental id. Each real instance has an id generated from its first timestamp.

More information about these variables can be obtained from the following publicly available documents:

• Option in Portuguese: R.E.V. Vargas. Base de dados e benchmarks para prognóstico de anomalias em sistemas de elevação de petróleo. Universidade Federal do Espírito Santo. Doctoral thesis. 2019. https://github.com/ricardovvargas/3w_dataset/raw/master/docs/doctoral_thesis_ricardo_vargas.pdf.
• Option in English: B.G. Carvalho. Evaluating machine learning techniques for detection of flow instability events in offshore oil wells. Universidade Federal do Espírito Santo. Master's degree dissertation. 2021. https://github.com/ricardovvargas/3w_dataset/raw/master/docs/master_degree_dissertation_bruno_carvalho.pdf.

Preprocessing

The following table shows the amount of instances that compose the 3W dataset, by knowledge source (real, simulated and hand-drawn instances) and by instance label.

class kmodel():
    
    def __init__(self, name, class_model, features, label, categories, batch_size, seq_length, 
                 train_df, val_df, test_df, tmp_path, reset_ts=False, verbose=1, class_bal=False):
        self.features = features
        self.n_features = len(features)
        self.label = label
        self.categories = categories
        self.n_classes = len(*categories)
        self.seq_length = seq_length
        self.batch_size = batch_size
        self.tmp_path = pathlib.Path(tmp_path)
        self.verbose = verbose
        self.class_bal = class_bal
        self.history = None
        self.class_rep_val = None
        self.class_rep_test = None
        self.class_rep_full = None
        self.conf_mat_val = None
        self.conf_mat_test = None
        self.conf_mat_full = None
        
        self.train = CustomDataGen(train_df, self.features, self.label, categories, 
                                   self.batch_size, self.seq_length, self.tmp_path)
        self.val = CustomDataGen(val_df, self.features, self.label, categories, 
                                 self.batch_size, self.seq_length, self.tmp_path)
        self.test = CustomDataGen(test_df, self.features, self.label, categories, 
                                  self.batch_size, self.seq_length, self.tmp_path)
        self.full = CustomDataGen(pd.concat([train_df, val_df, test_df], ignore_index=True), 
                                  self.features, self.label, categories, 
                                  self.batch_size, self.seq_length, self.tmp_path)
        
        if reset_ts:
            self.full.reset_ts()

        self.model = class_model()
        self.model.build(input_shape=(None, self.seq_length, 2*self.n_features))

        input = tf.keras.Input(shape=(self.seq_length, 2*self.n_features), name='input',
                              tensor=tf.zeros(shape=(self.batch_size, self.seq_length, 2*self.n_features)))
        y = self.model.call(input, training=True)
        
        #for layer in self.model.layers:
        #    print(self.model.name, layer.name, layer.input_shape, '->', layer.output_shape)
        #print('\n')
            
        if self.verbose > 0:
            print(self.model.summary())
        
        return
        
    def compile_and_fit(self, max_epochs=20, patience=3, lr=0.001):

        if self.class_bal:
            y_real = self.train.get_y()
            d = dict(zip(np.unique(y_real), compute_class_weight('balanced', classes=np.unique(y_real), y=y_real)))
            w = dict()
            for i, j in enumerate(categories[0]):
                if j in d.keys():
                    w[i] = d[j]
                else:
                    w[i] = 1.0
        else:
            w = None

        early_stopping = tf.keras.callbacks.EarlyStopping(monitor='val_f1_score',
                                                        patience=patience,
                                                        mode='max')

        checkpoint_filepath = self.tmp_path.joinpath('checkpoint')
        model_checkpoint_callback = tf.keras.callbacks.ModelCheckpoint(
            filepath=checkpoint_filepath,
            save_weights_only=True,
            monitor='val_f1_score',
            mode='max',
            save_best_only=True)

        f1_score = tfa.metrics.F1Score(num_classes=self.n_classes, average='macro', threshold=0.5)

        self.model.compile(loss=tf.keras.losses.CategoricalCrossentropy(),
                           optimizer=tf.keras.optimizers.Adam(learning_rate=lr),
                           metrics=[tf.keras.metrics.CategoricalAccuracy(), 
                                    tf.keras.metrics.Precision(),
                                    tf.keras.metrics.Recall(),
                                    f1_score])

        self.history = self.model.fit(self.train, epochs=max_epochs,
                                      validation_data=self.val, class_weight=w,
                                      callbacks=[early_stopping, model_checkpoint_callback])

        self.model.load_weights(checkpoint_filepath)
       
        self.class_rep_val, self.conf_mat_val = self.class_rep(self.val)
        
        self.class_rep_test, self.conf_mat_test = self.class_rep(self.test)

        self.class_rep_full, self.conf_mat_full = self.class_rep(self.full)

        if self.verbose > 0:
            self.plot_fithist('f1_score')
            
            self.print_class_rep(self.class_rep_val, 'Val Classification Report')
            self.plot_confusion_matrix(self.conf_mat_val, title='Confusion Matrix Val')
            
            self.print_class_rep(self.class_rep_test, 'Test Classification Report')
            self.plot_confusion_matrix(self.conf_mat_test, title='Confusion Matrix Test')

            self.print_class_rep(self.class_rep_full, 'All Data Classification Report')
            self.plot_confusion_matrix(self.conf_mat_full, title='Confusion Matrix Full')

        return

    # Plot training loss & metric 
    def plot_fithist(self, met):
        N = len(self.history.history['loss'])
        plt.style.use("ggplot")
        fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(10,5))
        fig.suptitle('Training Loss & '+met)

        ax1.plot(np.arange(0, N), self.history.history[met], label="train")
        ax1.plot(np.arange(0, N), self.history.history["val_"+met], label="val")
        ax1.set_title(met)
        ax1.set_xlabel("Epoch #")
        ax1.xaxis.get_major_locator().set_params(integer=True)
        ax1.set_ylabel(met)
        ax1.legend(loc="lower right")

        ax2.plot(np.arange(0, N), self.history.history["loss"], label="train")
        ax2.plot(np.arange(0, N), self.history.history["val_loss"], label="val")
        ax2.set_title("Loss")
        ax2.set_xlabel("Epoch #")
        ax2.xaxis.get_major_locator().set_params(integer=True)
        ax2.set_ylabel("Loss")
        ax2.legend(loc="upper right")

        plt.show()
        return
        
    def save(self):
        filename = self.model.name + '_' +\
            'batch_size_' + str(self.batch_size) + '-' +\
            'seq_length_' + str(self.seq_length) + '-' +\
            str(np.datetime64('now')) + '.h5'
        filename = filename.replace(':', '_')
        self.model.save_weights(filename)
        print('model weights saved in ' + filename)
        return
    
    def class_rep(self, dset):

        pred = self.model.predict(dset)

        y_pred = [self.categories[0][i] for i in np.argmax(pred, axis=1)]
        y_real = dset.get_y()

        return classification_report(y_real, y_pred, zero_division=0, output_dict=True),\
                                             confusion_matrix(y_real, y_pred, labels=self.categories[0])
    
    def print_class_rep(self, data_dict, title):
        # extracted from: https://github.com/scikit-learn/scikit-learn/blob/0fb307bf3/sklearn/metrics/_classification.py#L1825
        """Build a text report showing the main classification metrics.
        Read more in the :ref:`User Guide <classification_report>`.
        Parameters
        ----------
        report : string
            Text summary of the precision, recall, F1 score for each class.
            Dictionary returned if output_dict is True. Dictionary has the
            following structure::
                {'label 1': {'precision':0.5,
                             'recall':1.0,
                             'f1-score':0.67,
                             'support':1},
                 'label 2': { ... },
                  ...
                }
            The reported averages include macro average (averaging the unweighted
            mean per label), weighted average (averaging the support-weighted mean
            per label), and sample average (only for multilabel classification).
            Micro average (averaging the total true positives, false negatives and
            false positives) is only shown for multi-label or multi-class
            with a subset of classes, because it corresponds to accuracy otherwise.
            See also :func:`precision_recall_fscore_support` for more details
            on averages.
            Note that in binary classification, recall of the positive class
            is also known as "sensitivity"; recall of the negative class is
            "specificity".
        """

        non_label_keys = ["accuracy", "macro avg", "weighted avg"]
        y_type = "binary"
        digits = 2

        target_names = [
            "%s" % key for key in data_dict.keys() if key not in non_label_keys
        ]

        # labelled micro average
        micro_is_accuracy = (y_type == "multiclass" or y_type == "binary")

        headers = ["precision", "recall", "f1-score", "support"]
        p = [data_dict[l][headers[0]] for l in target_names]
        r = [data_dict[l][headers[1]] for l in target_names]
        f1 = [data_dict[l][headers[2]] for l in target_names]
        s = [data_dict[l][headers[3]] for l in target_names]

        rows = zip(target_names, p, r, f1, s)

        if y_type.startswith("multilabel"):
            average_options = ("micro", "macro", "weighted", "samples")
        else:
            average_options = ("micro", "macro", "weighted")

        longest_last_line_heading = "weighted avg"
        name_width = max(len(cn) for cn in target_names)
        width = max(name_width, len(longest_last_line_heading), digits)
        head_fmt = "{:>{width}s} " + " {:>9}" * len(headers)
        report = head_fmt.format("", *headers, width=width)
        report += "\n\n"
        row_fmt = "{:>{width}s} " + " {:>9.{digits}f}" * 3 + " {:>9}\n"
        for row in rows:
            report += row_fmt.format(*row, width=width, digits=digits)
        report += "\n"

        # compute all applicable averages
        for average in average_options:
            if average.startswith("micro") and micro_is_accuracy:
                line_heading = "accuracy"
            else:
                line_heading = average + " avg"

            if line_heading == "accuracy":
                avg = [data_dict[line_heading], sum(s)]
                row_fmt_accuracy = "{:>{width}s} " + \
                        " {:>9.{digits}}" * 2 + " {:>9.{digits}f}" + \
                        " {:>9}\n"
                report += row_fmt_accuracy.format(line_heading, "", "",
                                                  *avg, width=width,
                                                  digits=digits)
            else:
                avg = list(data_dict[line_heading].values())
                report += row_fmt.format(line_heading, *avg,
                                         width=width, digits=digits)
        print(title)
        print(report)
        
        return

    def plot_confusion_matrix(self, cm, title='Confusion Matrix'):
        con = np.zeros((self.n_classes, self.n_classes))
        for x in range(self.n_classes):
            for y in range(self.n_classes):
                if np.sum(cm[x,:]) > 0:
                    con[x,y] = cm[x,y]/np.sum(cm[x,:])

        plt.figure(figsize=(10,6))
        sns.set(font_scale=0.75) # for label size
        ax = sns.heatmap(con, annot=True,fmt='.3f', cmap='Blues',xticklabels= self.categories[0] , 
                         yticklabels= self.categories[0], cbar=False)
        plt.tight_layout()
        ax.set_xlabel('Predicted', fontsize=16)
        ax.set_ylabel('Real', fontsize=16)
        ax.set_title(title, fontsize=18)
        plt.show()

class kmodel_from_file(kmodel):
    def __init__(self, model, input_shape, filename):

        self.model = model()
        self.model.build(input_shape=(None,)+input_shape)
        self.model.load_weights(filename)
        self.model.summary()
        
    def class_rep(self, dset):
        self.categories = dset.categories
        self.n_classes = len(*self.categories)
        cr, cm = super().class_rep(dset)
        super().print_class_rep(cr, 'Classification Report')
        super().plot_confusion_matrix(cm, title='Confusion Matrix')
        return

class param_fit():
    def __init__(self, name, class_model, features, label, categories, batch_sizes, seq_lengths, learning_rates, 
                 train_df, test_df, tmp_path):
        self.name = name 
        self.class_model = class_model
        self.features = features
        self.label = label
        self.categories = categories
        self.batch_sizes = batch_sizes
        self.seq_lengths = seq_lengths
        self.learning_rates = learning_rates
        self.train_df = train_df
        self.test_df = test_df
        self.tmp_path = tmp_path
        self.report = dict()
        return
    
    def fit(self, n_splits=5, max_epochs=50, patience=3):
        kf = sklearn.model_selection.StratifiedKFold(n_splits=n_splits)
        
        self.report['batch_size'] = []
        self.report['seq_length'] = []
        self.report['learning_rate'] = []
        self.report['cv_n'] = []
        self.report['class_rep_val'] = []
        self.report['class_rep_test'] = []
        for batch_size in self.batch_sizes:
            breset = True
            for seq_length in self.seq_lengths:
                breset = True
                for learning_rate in self.learning_rates:
                    for i, (train_index, val_index) in enumerate(kf.split(self.train_df, self.train_df['label'])):
                        print('\n', i, batch_size, seq_length, learning_rate,
                              len(train_index), len(val_index), '\n')
                        model = kmodel(self.name, self.class_model, self.features, self.label, self.categories, 
                                   batch_size, seq_length, 
                                   self.train_df.iloc[train_index].reset_index(drop=True), 
                                   self.train_df.iloc[val_index].reset_index(drop=True), 
                                   self.test_df, 
                                   'D:/datatmp', reset_ts=breset, verbose=0)
                        model.compile_and_fit(max_epochs=max_epochs, patience=patience, lr=learning_rate)
                        breset = False
                        self.report['batch_size'].append(batch_size)
                        self.report['seq_length'].append(seq_length)
                        self.report['learning_rate'].append(learning_rate)
                        self.report['cv_n'].append(i)
                        self.report['class_rep_val'].append(model.class_rep_val)
                        self.report['class_rep_test'].append(model.class_rep_test)
                        
        
        rep_df = pd.DataFrame(self.report)
        rep_df['val_accuracy'] = rep_df.class_rep_val.apply(pd.Series)['accuracy']
        rep_df['val_precision'] = rep_df.class_rep_val.apply(pd.Series)['macro avg'].apply(pd.Series)['precision']
        rep_df['val_recall'] = rep_df.class_rep_val.apply(pd.Series)['macro avg'].apply(pd.Series)['recall']
        rep_df['val_f1_score'] = rep_df.class_rep_val.apply(pd.Series)['macro avg'].apply(pd.Series)['f1-score']
        rep_df.drop(['class_rep_val', 'class_rep_test'], axis=1, inplace=True)
        
        fdict = {'val_accuracy': ['mean', 'std'], 'val_precision': ['mean', 'std'],
                 'val_recall': ['mean', 'std'], 'val_f1_score': ['mean', 'std']}
        display(rep_df.groupby(['batch_size', 'seq_length', 'learning_rate']).agg(fdict))        
        
        filename = self.name + '_report_' +\
            str(np.datetime64('now')) + '.pkl'
        filename = filename.replace(':', '_')

        with open(filename, 'wb') as f:
            pickle.dump(self.report, f)

Linear model

class LinearModel(tf.keras.Model):

  def __init__(self):
    super().__init__()
    self.flatten = layers.Flatten()
    self.norm = layers.LayerNormalization()
    self.dense = layers.Dense(units=17, activation='sigmoid')

  def call(self, inputs, training=False):
    x = self.flatten(inputs)
    x = self.norm(x)
    return self.dense(x)

%%time

train_df, test_df = dset.split(real=True, simul=True, drawn=True, test_size=0.2, val_size=0, sample_n=300)

tf.keras.backend.clear_session()
tf.compat.v1.reset_default_graph()

linear_param_fit = param_fit('linear_param_fit', LinearModel, flist0, 'class', categories, 
                             [32, 64], [15, 30], [0.0001, 0.001], 
                             train_df, test_df, 'D:/datatmp')
linear_param_fit.fit(n_splits=5, max_epochs=50)

Instances Train: 237  Test: 60

 0 32 15 0.0001 189 48 

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

Epoch 1/50
2636/2636 [==============================] - 67s 25ms/step - loss: 2.3815 - categorical_accuracy: 0.3326 - precision: 0.0927 - recall: 0.7150 - f1_score: 0.1292 - val_loss: 2.0120 - val_categorical_accuracy: 0.4345 - val_precision: 0.1036 - val_recall: 0.7958 - val_f1_score: 0.1330
Epoch 2/50
2636/2636 [==============================] - 38s 14ms/step - loss: 1.7201 - categorical_accuracy: 0.4939 - precision: 0.1155 - recall: 0.8580 - f1_score: 0.1466 - val_loss: 1.7653 - val_categorical_accuracy: 0.4907 - val_precision: 0.1170 - val_recall: 0.8440 - val_f1_score: 0.1369
Epoch 3/50
2636/2636 [==============================] - 37s 14ms/step - loss: 1.5257 - categorical_accuracy: 0.5473 - precision: 0.1237 - recall: 0.8933 - f1_score: 0.1528 - val_loss: 1.6732 - val_categorical_accuracy: 0.5202 - val_precision: 0.1208 - val_recall: 0.8710 - val_f1_score: 0.1424
Epoch 4/50
2636/2636 [==============================] - 37s 14ms/step - loss: 1.4371 - categorical_accuracy: 0.5720 - precision: 0.1259 - recall: 0.9112 - f1_score: 0.1580 - val_loss: 1.6255 - val_categorical_accuracy: 0.5412 - val_precision: 0.1231 - val_recall: 0.8887 - val_f1_score: 0.1491
Epoch 5/50
2636/2636 [==============================] - 37s 14ms/step - loss: 1.3870 - categorical_accuracy: 0.5850 - precision: 0.1275 - recall: 0.9255 - f1_score: 0.1650 - val_loss: 1.6072 - val_categorical_accuracy: 0.5417 - val_precision: 0.1235 - val_recall: 0.9025 - val_f1_score: 0.1535
Epoch 6/50
2636/2636 [==============================] - 38s 14ms/step - loss: 1.3574 - categorical_accuracy: 0.5925 - precision: 0.1285 - recall: 0.9343 - f1_score: 0.1688 - val_loss: 1.5982 - val_categorical_accuracy: 0.5502 - val_precision: 0.1242 - val_recall: 0.9117 - val_f1_score: 0.1562
Epoch 7/50
2636/2636 [==============================] - 34s 13ms/step - loss: 1.3354 - categorical_accuracy: 0.5982 - precision: 0.1289 - recall: 0.9383 - f1_score: 0.1722 - val_loss: 1.5940 - val_categorical_accuracy: 0.5519 - val_precision: 0.1244 - val_recall: 0.9186 - val_f1_score: 0.1590
Epoch 8/50
2636/2636 [==============================] - 36s 14ms/step - loss: 1.3206 - categorical_accuracy: 0.6041 - precision: 0.1283 - recall: 0.9427 - f1_score: 0.1777 - val_loss: 1.5924 - val_categorical_accuracy: 0.5612 - val_precision: 0.1241 - val_recall: 0.9200 - val_f1_score: 0.1608
Epoch 9/50
2636/2636 [==============================] - 34s 13ms/step - loss: 1.3077 - categorical_accuracy: 0.6071 - precision: 0.1279 - recall: 0.9435 - f1_score: 0.1901 - val_loss: 1.5887 - val_categorical_accuracy: 0.5598 - val_precision: 0.1247 - val_recall: 0.9240 - val_f1_score: 0.1634
Epoch 10/50
2636/2636 [==============================] - 34s 13ms/step - loss: 1.2969 - categorical_accuracy: 0.6107 - precision: 0.1286 - recall: 0.9481 - f1_score: 0.1946 - val_loss: 1.5967 - val_categorical_accuracy: 0.5623 - val_precision: 0.1251 - val_recall: 0.9268 - val_f1_score: 0.1655
Epoch 11/50
2636/2636 [==============================] - 36s 13ms/step - loss: 1.2900 - categorical_accuracy: 0.6132 - precision: 0.1281 - recall: 0.9479 - f1_score: 0.1963 - val_loss: 1.5925 - val_categorical_accuracy: 0.5646 - val_precision: 0.1245 - val_recall: 0.9283 - val_f1_score: 0.1670
Epoch 12/50
2636/2636 [==============================] - 38s 14ms/step - loss: 1.2828 - categorical_accuracy: 0.6159 - precision: 0.1287 - recall: 0.9491 - f1_score: 0.1996 - val_loss: 1.5990 - val_categorical_accuracy: 0.5713 - val_precision: 0.1242 - val_recall: 0.9295 - val_f1_score: 0.1689
Epoch 13/50
2636/2636 [==============================] - 35s 13ms/step - loss: 1.2766 - categorical_accuracy: 0.6178 - precision: 0.1283 - recall: 0.9513 - f1_score: 0.2008 - val_loss: 1.5994 - val_categorical_accuracy: 0.5674 - val_precision: 0.1246 - val_recall: 0.9314 - val_f1_score: 0.1712
Epoch 14/50
2636/2636 [==============================] - 35s 13ms/step - loss: 1.2715 - categorical_accuracy: 0.6168 - precision: 0.1284 - recall: 0.9542 - f1_score: 0.2047 - val_loss: 1.6064 - val_categorical_accuracy: 0.5711 - val_precision: 0.1240 - val_recall: 0.9312 - val_f1_score: 0.1683
Epoch 15/50
2636/2636 [==============================] - 36s 13ms/step - loss: 1.2670 - categorical_accuracy: 0.6241 - precision: 0.1282 - recall: 0.9527 - f1_score: 0.2033 - val_loss: 1.6004 - val_categorical_accuracy: 0.5709 - val_precision: 0.1239 - val_recall: 0.9319 - val_f1_score: 0.1697
Epoch 16/50
2636/2636 [==============================] - 38s 14ms/step - loss: 1.2627 - categorical_accuracy: 0.6227 - precision: 0.1281 - recall: 0.9541 - f1_score: 0.2028 - val_loss: 1.6151 - val_categorical_accuracy: 0.5712 - val_precision: 0.1238 - val_recall: 0.9309 - val_f1_score: 0.1711

 1 32 15 0.0001 189 48 

Epoch 1/50
2643/2643 [==============================] - 40s 15ms/step - loss: 2.3237 - categorical_accuracy: 0.3533 - precision_1: 0.0969 - recall_1: 0.7423 - f1_score: 0.1342 - val_loss: 1.9285 - val_categorical_accuracy: 0.4347 - val_precision_1: 0.1125 - val_recall_1: 0.8213 - val_f1_score: 0.1478
Epoch 2/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.7316 - categorical_accuracy: 0.4959 - precision_1: 0.1203 - recall_1: 0.8641 - f1_score: 0.1528 - val_loss: 1.6516 - val_categorical_accuracy: 0.4983 - val_precision_1: 0.1224 - val_recall_1: 0.8683 - val_f1_score: 0.1561
Epoch 3/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.5417 - categorical_accuracy: 0.5460 - precision_1: 0.1279 - recall_1: 0.8963 - f1_score: 0.1626 - val_loss: 1.5326 - val_categorical_accuracy: 0.5258 - val_precision_1: 0.1279 - val_recall_1: 0.8975 - val_f1_score: 0.1631
Epoch 4/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.4535 - categorical_accuracy: 0.5672 - precision_1: 0.1304 - recall_1: 0.9154 - f1_score: 0.1702 - val_loss: 1.4740 - val_categorical_accuracy: 0.5510 - val_precision_1: 0.1285 - val_recall_1: 0.9091 - val_f1_score: 0.1701
Epoch 5/50
2643/2643 [==============================] - 40s 15ms/step - loss: 1.4066 - categorical_accuracy: 0.5803 - precision_1: 0.1321 - recall_1: 0.9251 - f1_score: 0.1762 - val_loss: 1.4341 - val_categorical_accuracy: 0.5586 - val_precision_1: 0.1297 - val_recall_1: 0.9224 - val_f1_score: 0.1742
Epoch 6/50
2643/2643 [==============================] - 40s 15ms/step - loss: 1.3761 - categorical_accuracy: 0.5891 - precision_1: 0.1313 - recall_1: 0.9331 - f1_score: 0.1772 - val_loss: 1.4058 - val_categorical_accuracy: 0.5697 - val_precision_1: 0.1312 - val_recall_1: 0.9281 - val_f1_score: 0.1808
Epoch 7/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.3565 - categorical_accuracy: 0.5987 - precision_1: 0.1319 - recall_1: 0.9349 - f1_score: 0.1834 - val_loss: 1.3932 - val_categorical_accuracy: 0.5715 - val_precision_1: 0.1314 - val_recall_1: 0.9333 - val_f1_score: 0.1857
Epoch 8/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.3410 - categorical_accuracy: 0.6027 - precision_1: 0.1316 - recall_1: 0.9379 - f1_score: 0.1879 - val_loss: 1.3837 - val_categorical_accuracy: 0.5725 - val_precision_1: 0.1309 - val_recall_1: 0.9393 - val_f1_score: 0.1908
Epoch 9/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.3294 - categorical_accuracy: 0.6050 - precision_1: 0.1313 - recall_1: 0.9400 - f1_score: 0.1901 - val_loss: 1.3671 - val_categorical_accuracy: 0.5860 - val_precision_1: 0.1317 - val_recall_1: 0.9417 - val_f1_score: 0.1938
Epoch 10/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.3198 - categorical_accuracy: 0.6123 - precision_1: 0.1317 - recall_1: 0.9414 - f1_score: 0.1953 - val_loss: 1.3612 - val_categorical_accuracy: 0.5833 - val_precision_1: 0.1316 - val_recall_1: 0.9441 - val_f1_score: 0.1979
Epoch 11/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.3114 - categorical_accuracy: 0.6138 - precision_1: 0.1315 - recall_1: 0.9423 - f1_score: 0.1984 - val_loss: 1.3536 - val_categorical_accuracy: 0.5872 - val_precision_1: 0.1324 - val_recall_1: 0.9439 - val_f1_score: 0.1979
Epoch 12/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.3049 - categorical_accuracy: 0.6155 - precision_1: 0.1316 - recall_1: 0.9433 - f1_score: 0.2004 - val_loss: 1.3461 - val_categorical_accuracy: 0.5956 - val_precision_1: 0.1330 - val_recall_1: 0.9479 - val_f1_score: 0.1996
Epoch 13/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.2993 - categorical_accuracy: 0.6191 - precision_1: 0.1316 - recall_1: 0.9450 - f1_score: 0.2107 - val_loss: 1.3434 - val_categorical_accuracy: 0.5985 - val_precision_1: 0.1324 - val_recall_1: 0.9470 - val_f1_score: 0.1965
Epoch 14/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.2939 - categorical_accuracy: 0.6196 - precision_1: 0.1315 - recall_1: 0.9465 - f1_score: 0.2120 - val_loss: 1.3467 - val_categorical_accuracy: 0.5970 - val_precision_1: 0.1321 - val_recall_1: 0.9447 - val_f1_score: 0.1967
Epoch 15/50
2643/2643 [==============================] - 39s 15ms/step - loss: 1.2904 - categorical_accuracy: 0.6204 - precision_1: 0.1314 - recall_1: 0.9457 - f1_score: 0.2089 - val_loss: 1.3481 - val_categorical_accuracy: 0.5973 - val_precision_1: 0.1341 - val_recall_1: 0.9493 - val_f1_score: 0.2015
Epoch 16/50
2643/2643 [==============================] - 37s 14ms/step - loss: 1.2873 - categorical_accuracy: 0.6218 - precision_1: 0.1321 - recall_1: 0.9474 - f1_score: 0.2218 - val_loss: 1.3448 - val_categorical_accuracy: 0.5925 - val_precision_1: 0.1337 - val_recall_1: 0.9520 - val_f1_score: 0.2004
Epoch 17/50
2643/2643 [==============================] - 36s 14ms/step - loss: 1.2827 - categorical_accuracy: 0.6257 - precision_1: 0.1324 - recall_1: 0.9474 - f1_score: 0.2211 - val_loss: 1.3332 - val_categorical_accuracy: 0.5996 - val_precision_1: 0.1328 - val_recall_1: 0.9561 - val_f1_score: 0.1991
Epoch 18/50
2643/2643 [==============================] - 36s 14ms/step - loss: 1.2801 - categorical_accuracy: 0.6227 - precision_1: 0.1318 - recall_1: 0.9488 - f1_score: 0.2263 - val_loss: 1.3315 - val_categorical_accuracy: 0.6069 - val_precision_1: 0.1331 - val_recall_1: 0.9540 - val_f1_score: 0.1990

 2 32 15 0.0001 190 47 

Epoch 1/50
2512/2512 [==============================] - 38s 15ms/step - loss: 2.3215 - categorical_accuracy: 0.3605 - precision_2: 0.1058 - recall_2: 0.7259 - f1_score: 0.1380 - val_loss: 2.1867 - val_categorical_accuracy: 0.3474 - val_precision_2: 0.1092 - val_recall_2: 0.7627 - val_f1_score: 0.1481
Epoch 2/50
2512/2512 [==============================] - 36s 14ms/step - loss: 1.7180 - categorical_accuracy: 0.5087 - precision_2: 0.1303 - recall_2: 0.8434 - f1_score: 0.1558 - val_loss: 2.0351 - val_categorical_accuracy: 0.4213 - val_precision_2: 0.1197 - val_recall_2: 0.8092 - val_f1_score: 0.1512
Epoch 3/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.5233 - categorical_accuracy: 0.5557 - precision_2: 0.1404 - recall_2: 0.8801 - f1_score: 0.1627 - val_loss: 2.0132 - val_categorical_accuracy: 0.4407 - val_precision_2: 0.1292 - val_recall_2: 0.8300 - val_f1_score: 0.1553
Epoch 4/50
2512/2512 [==============================] - 39s 16ms/step - loss: 1.4320 - categorical_accuracy: 0.5767 - precision_2: 0.1445 - recall_2: 0.8993 - f1_score: 0.1711 - val_loss: 2.0195 - val_categorical_accuracy: 0.4493 - val_precision_2: 0.1293 - val_recall_2: 0.8455 - val_f1_score: 0.1576
Epoch 5/50
2512/2512 [==============================] - 35s 14ms/step - loss: 1.3809 - categorical_accuracy: 0.5854 - precision_2: 0.1427 - recall_2: 0.9158 - f1_score: 0.1778 - val_loss: 2.0388 - val_categorical_accuracy: 0.4501 - val_precision_2: 0.1311 - val_recall_2: 0.8597 - val_f1_score: 0.1614
Epoch 6/50
2512/2512 [==============================] - 34s 14ms/step - loss: 1.3504 - categorical_accuracy: 0.5930 - precision_2: 0.1431 - recall_2: 0.9291 - f1_score: 0.1878 - val_loss: 2.0476 - val_categorical_accuracy: 0.4481 - val_precision_2: 0.1308 - val_recall_2: 0.8654 - val_f1_score: 0.1636
Epoch 7/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.3281 - categorical_accuracy: 0.5954 - precision_2: 0.1427 - recall_2: 0.9347 - f1_score: 0.1934 - val_loss: 2.0732 - val_categorical_accuracy: 0.4622 - val_precision_2: 0.1282 - val_recall_2: 0.8680 - val_f1_score: 0.1617
Epoch 8/50
2512/2512 [==============================] - 36s 14ms/step - loss: 1.3116 - categorical_accuracy: 0.6054 - precision_2: 0.1417 - recall_2: 0.9390 - f1_score: 0.1992 - val_loss: 2.1081 - val_categorical_accuracy: 0.4592 - val_precision_2: 0.1273 - val_recall_2: 0.8712 - val_f1_score: 0.1611
Epoch 9/50
2512/2512 [==============================] - 36s 14ms/step - loss: 1.2999 - categorical_accuracy: 0.6060 - precision_2: 0.1410 - recall_2: 0.9427 - f1_score: 0.2026 - val_loss: 2.1299 - val_categorical_accuracy: 0.4626 - val_precision_2: 0.1264 - val_recall_2: 0.8723 - val_f1_score: 0.1616

 3 32 15 0.0001 190 47 

Epoch 1/50
2678/2678 [==============================] - 37s 13ms/step - loss: 2.3113 - categorical_accuracy: 0.3426 - precision_3: 0.0989 - recall_3: 0.7614 - f1_score: 0.1403 - val_loss: 1.8747 - val_categorical_accuracy: 0.4779 - val_precision_3: 0.1108 - val_recall_3: 0.8228 - val_f1_score: 0.1435
Epoch 2/50
2678/2678 [==============================] - 34s 13ms/step - loss: 1.7457 - categorical_accuracy: 0.5058 - precision_3: 0.1144 - recall_3: 0.8578 - f1_score: 0.1497 - val_loss: 1.6260 - val_categorical_accuracy: 0.5439 - val_precision_3: 0.1198 - val_recall_3: 0.8687 - val_f1_score: 0.1469
Epoch 3/50
2678/2678 [==============================] - 38s 14ms/step - loss: 1.5587 - categorical_accuracy: 0.5511 - precision_3: 0.1232 - recall_3: 0.8901 - f1_score: 0.1554 - val_loss: 1.5118 - val_categorical_accuracy: 0.5673 - val_precision_3: 0.1250 - val_recall_3: 0.8961 - val_f1_score: 0.1534
Epoch 4/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.4649 - categorical_accuracy: 0.5734 - precision_3: 0.1269 - recall_3: 0.9123 - f1_score: 0.1614 - val_loss: 1.4626 - val_categorical_accuracy: 0.5765 - val_precision_3: 0.1266 - val_recall_3: 0.9106 - val_f1_score: 0.1584
Epoch 5/50
2678/2678 [==============================] - 35s 13ms/step - loss: 1.4102 - categorical_accuracy: 0.5821 - precision_3: 0.1282 - recall_3: 0.9258 - f1_score: 0.1668 - val_loss: 1.4252 - val_categorical_accuracy: 0.5884 - val_precision_3: 0.1281 - val_recall_3: 0.9242 - val_f1_score: 0.1659
Epoch 6/50
2678/2678 [==============================] - 38s 14ms/step - loss: 1.3806 - categorical_accuracy: 0.5901 - precision_3: 0.1293 - recall_3: 0.9368 - f1_score: 0.1775 - val_loss: 1.4033 - val_categorical_accuracy: 0.5984 - val_precision_3: 0.1289 - val_recall_3: 0.9305 - val_f1_score: 0.1694
Epoch 7/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3577 - categorical_accuracy: 0.5957 - precision_3: 0.1296 - recall_3: 0.9399 - f1_score: 0.1845 - val_loss: 1.3915 - val_categorical_accuracy: 0.6071 - val_precision_3: 0.1291 - val_recall_3: 0.9339 - val_f1_score: 0.1731
Epoch 8/50
2678/2678 [==============================] - 36s 13ms/step - loss: 1.3404 - categorical_accuracy: 0.5984 - precision_3: 0.1305 - recall_3: 0.9472 - f1_score: 0.1896 - val_loss: 1.3898 - val_categorical_accuracy: 0.6094 - val_precision_3: 0.1283 - val_recall_3: 0.9348 - val_f1_score: 0.1742
Epoch 9/50
2678/2678 [==============================] - 36s 13ms/step - loss: 1.3281 - categorical_accuracy: 0.6017 - precision_3: 0.1299 - recall_3: 0.9509 - f1_score: 0.1928 - val_loss: 1.3802 - val_categorical_accuracy: 0.6149 - val_precision_3: 0.1286 - val_recall_3: 0.9349 - val_f1_score: 0.1756
Epoch 10/50
2678/2678 [==============================] - 35s 13ms/step - loss: 1.3168 - categorical_accuracy: 0.6060 - precision_3: 0.1304 - recall_3: 0.9508 - f1_score: 0.2000 - val_loss: 1.3862 - val_categorical_accuracy: 0.5973 - val_precision_3: 0.1276 - val_recall_3: 0.9365 - val_f1_score: 0.1765
Epoch 11/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3105 - categorical_accuracy: 0.6085 - precision_3: 0.1296 - recall_3: 0.9545 - f1_score: 0.1989 - val_loss: 1.3797 - val_categorical_accuracy: 0.6041 - val_precision_3: 0.1288 - val_recall_3: 0.9374 - val_f1_score: 0.1797
Epoch 12/50
2678/2678 [==============================] - 36s 14ms/step - loss: 1.3026 - categorical_accuracy: 0.6145 - precision_3: 0.1305 - recall_3: 0.9549 - f1_score: 0.2041 - val_loss: 1.3720 - val_categorical_accuracy: 0.6102 - val_precision_3: 0.1286 - val_recall_3: 0.9394 - val_f1_score: 0.1826
Epoch 13/50
2678/2678 [==============================] - 36s 13ms/step - loss: 1.2975 - categorical_accuracy: 0.6105 - precision_3: 0.1304 - recall_3: 0.9559 - f1_score: 0.2057 - val_loss: 1.3590 - val_categorical_accuracy: 0.6209 - val_precision_3: 0.1282 - val_recall_3: 0.9396 - val_f1_score: 0.1824
Epoch 14/50
2678/2678 [==============================] - 39s 14ms/step - loss: 1.2923 - categorical_accuracy: 0.6169 - precision_3: 0.1299 - recall_3: 0.9567 - f1_score: 0.2068 - val_loss: 1.3622 - val_categorical_accuracy: 0.6148 - val_precision_3: 0.1289 - val_recall_3: 0.9401 - val_f1_score: 0.1845
Epoch 15/50
2678/2678 [==============================] - 36s 13ms/step - loss: 1.2877 - categorical_accuracy: 0.6164 - precision_3: 0.1300 - recall_3: 0.9579 - f1_score: 0.2104 - val_loss: 1.3626 - val_categorical_accuracy: 0.6193 - val_precision_3: 0.1284 - val_recall_3: 0.9407 - val_f1_score: 0.1848
Epoch 16/50
2678/2678 [==============================] - 35s 13ms/step - loss: 1.2844 - categorical_accuracy: 0.6153 - precision_3: 0.1302 - recall_3: 0.9586 - f1_score: 0.2144 - val_loss: 1.3566 - val_categorical_accuracy: 0.6202 - val_precision_3: 0.1276 - val_recall_3: 0.9410 - val_f1_score: 0.1860
Epoch 17/50
2678/2678 [==============================] - 36s 14ms/step - loss: 1.2805 - categorical_accuracy: 0.6184 - precision_3: 0.1301 - recall_3: 0.9579 - f1_score: 0.2160 - val_loss: 1.3535 - val_categorical_accuracy: 0.6272 - val_precision_3: 0.1279 - val_recall_3: 0.9406 - val_f1_score: 0.1855
Epoch 18/50
2678/2678 [==============================] - 36s 14ms/step - loss: 1.2770 - categorical_accuracy: 0.6206 - precision_3: 0.1300 - recall_3: 0.9591 - f1_score: 0.2179 - val_loss: 1.3495 - val_categorical_accuracy: 0.6273 - val_precision_3: 0.1279 - val_recall_3: 0.9414 - val_f1_score: 0.1871
Epoch 19/50
2678/2678 [==============================] - 36s 13ms/step - loss: 1.2730 - categorical_accuracy: 0.6239 - precision_3: 0.1299 - recall_3: 0.9606 - f1_score: 0.2195 - val_loss: 1.3590 - val_categorical_accuracy: 0.6203 - val_precision_3: 0.1281 - val_recall_3: 0.9420 - val_f1_score: 0.1884
Epoch 20/50
2678/2678 [==============================] - 36s 13ms/step - loss: 1.2711 - categorical_accuracy: 0.6275 - precision_3: 0.1299 - recall_3: 0.9590 - f1_score: 0.2211 - val_loss: 1.3524 - val_categorical_accuracy: 0.6211 - val_precision_3: 0.1278 - val_recall_3: 0.9427 - val_f1_score: 0.1874
Epoch 21/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.2692 - categorical_accuracy: 0.6220 - precision_3: 0.1303 - recall_3: 0.9604 - f1_score: 0.2203 - val_loss: 1.3525 - val_categorical_accuracy: 0.6202 - val_precision_3: 0.1273 - val_recall_3: 0.9417 - val_f1_score: 0.1879
Epoch 22/50
2678/2678 [==============================] - 38s 14ms/step - loss: 1.2641 - categorical_accuracy: 0.6277 - precision_3: 0.1299 - recall_3: 0.9621 - f1_score: 0.2231 - val_loss: 1.3638 - val_categorical_accuracy: 0.6085 - val_precision_3: 0.1279 - val_recall_3: 0.9426 - val_f1_score: 0.1873

 4 32 15 0.0001 190 47 

Epoch 1/50
2679/2679 [==============================] - 44s 16ms/step - loss: 2.3235 - categorical_accuracy: 0.3493 - precision_4: 0.1024 - recall_4: 0.7580 - f1_score: 0.1444 - val_loss: 1.8035 - val_categorical_accuracy: 0.4754 - val_precision_4: 0.1232 - val_recall_4: 0.8676 - val_f1_score: 0.1528
Epoch 2/50
2679/2679 [==============================] - 37s 14ms/step - loss: 1.7599 - categorical_accuracy: 0.4851 - precision_4: 0.1230 - recall_4: 0.8642 - f1_score: 0.1570 - val_loss: 1.5527 - val_categorical_accuracy: 0.5452 - val_precision_4: 0.1338 - val_recall_4: 0.9052 - val_f1_score: 0.1553
Epoch 3/50
2679/2679 [==============================] - 36s 14ms/step - loss: 1.5688 - categorical_accuracy: 0.5395 - precision_4: 0.1332 - recall_4: 0.8984 - f1_score: 0.1650 - val_loss: 1.4485 - val_categorical_accuracy: 0.5708 - val_precision_4: 0.1380 - val_recall_4: 0.9268 - val_f1_score: 0.1623
Epoch 4/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.4790 - categorical_accuracy: 0.5609 - precision_4: 0.1344 - recall_4: 0.9203 - f1_score: 0.1746 - val_loss: 1.4054 - val_categorical_accuracy: 0.5811 - val_precision_4: 0.1351 - val_recall_4: 0.9297 - val_f1_score: 0.1694
Epoch 5/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.4292 - categorical_accuracy: 0.5731 - precision_4: 0.1344 - recall_4: 0.9296 - f1_score: 0.1816 - val_loss: 1.3824 - val_categorical_accuracy: 0.5903 - val_precision_4: 0.1349 - val_recall_4: 0.9348 - val_f1_score: 0.1721
Epoch 6/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3994 - categorical_accuracy: 0.5834 - precision_4: 0.1343 - recall_4: 0.9368 - f1_score: 0.1884 - val_loss: 1.3564 - val_categorical_accuracy: 0.5966 - val_precision_4: 0.1341 - val_recall_4: 0.9405 - val_f1_score: 0.1754
Epoch 7/50
2679/2679 [==============================] - 39s 14ms/step - loss: 1.3757 - categorical_accuracy: 0.5890 - precision_4: 0.1340 - recall_4: 0.9431 - f1_score: 0.1848 - val_loss: 1.3402 - val_categorical_accuracy: 0.6027 - val_precision_4: 0.1354 - val_recall_4: 0.9404 - val_f1_score: 0.1763
Epoch 8/50
2679/2679 [==============================] - 40s 15ms/step - loss: 1.3600 - categorical_accuracy: 0.5912 - precision_4: 0.1347 - recall_4: 0.9453 - f1_score: 0.1960 - val_loss: 1.3342 - val_categorical_accuracy: 0.6048 - val_precision_4: 0.1350 - val_recall_4: 0.9474 - val_f1_score: 0.1781
Epoch 9/50
2679/2679 [==============================] - 40s 15ms/step - loss: 1.3471 - categorical_accuracy: 0.5986 - precision_4: 0.1349 - recall_4: 0.9474 - f1_score: 0.2066 - val_loss: 1.3227 - val_categorical_accuracy: 0.6069 - val_precision_4: 0.1340 - val_recall_4: 0.9499 - val_f1_score: 0.1785
Epoch 10/50
2679/2679 [==============================] - 39s 15ms/step - loss: 1.3367 - categorical_accuracy: 0.6031 - precision_4: 0.1344 - recall_4: 0.9493 - f1_score: 0.2031 - val_loss: 1.3193 - val_categorical_accuracy: 0.6114 - val_precision_4: 0.1330 - val_recall_4: 0.9448 - val_f1_score: 0.1781
Epoch 11/50
2679/2679 [==============================] - 34s 13ms/step - loss: 1.3281 - categorical_accuracy: 0.6036 - precision_4: 0.1337 - recall_4: 0.9512 - f1_score: 0.2132 - val_loss: 1.3106 - val_categorical_accuracy: 0.6213 - val_precision_4: 0.1343 - val_recall_4: 0.9447 - val_f1_score: 0.1808
Epoch 12/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3222 - categorical_accuracy: 0.6073 - precision_4: 0.1344 - recall_4: 0.9511 - f1_score: 0.2070 - val_loss: 1.3039 - val_categorical_accuracy: 0.6178 - val_precision_4: 0.1342 - val_recall_4: 0.9472 - val_f1_score: 0.1845
Epoch 13/50
2679/2679 [==============================] - 37s 14ms/step - loss: 1.3153 - categorical_accuracy: 0.6098 - precision_4: 0.1341 - recall_4: 0.9523 - f1_score: 0.2172 - val_loss: 1.3017 - val_categorical_accuracy: 0.6184 - val_precision_4: 0.1332 - val_recall_4: 0.9507 - val_f1_score: 0.1867
Epoch 14/50
2679/2679 [==============================] - 35s 13ms/step - loss: 1.3102 - categorical_accuracy: 0.6126 - precision_4: 0.1337 - recall_4: 0.9520 - f1_score: 0.2128 - val_loss: 1.2943 - val_categorical_accuracy: 0.6276 - val_precision_4: 0.1349 - val_recall_4: 0.9508 - val_f1_score: 0.1880
Epoch 15/50
2679/2679 [==============================] - 43s 16ms/step - loss: 1.3049 - categorical_accuracy: 0.6120 - precision_4: 0.1339 - recall_4: 0.9532 - f1_score: 0.2226 - val_loss: 1.3022 - val_categorical_accuracy: 0.6224 - val_precision_4: 0.1338 - val_recall_4: 0.9537 - val_f1_score: 0.1873
Epoch 16/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3021 - categorical_accuracy: 0.6172 - precision_4: 0.1344 - recall_4: 0.9542 - f1_score: 0.2267 - val_loss: 1.2889 - val_categorical_accuracy: 0.6250 - val_precision_4: 0.1347 - val_recall_4: 0.9538 - val_f1_score: 0.1878
Epoch 17/50
2679/2679 [==============================] - 37s 14ms/step - loss: 1.2981 - categorical_accuracy: 0.6183 - precision_4: 0.1340 - recall_4: 0.9544 - f1_score: 0.2301 - val_loss: 1.2904 - val_categorical_accuracy: 0.6200 - val_precision_4: 0.1336 - val_recall_4: 0.9502 - val_f1_score: 0.1872

 0 32 15 0.001 189 48 

Epoch 1/50

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

2636/2636 [==============================] - 42s 16ms/step - loss: 1.6167 - categorical_accuracy: 0.5376 - precision_5: 0.1241 - recall_5: 0.8977 - f1_score: 0.1576 - val_loss: 1.5686 - val_categorical_accuracy: 0.5692 - val_precision_5: 0.1290 - val_recall_5: 0.9283 - val_f1_score: 0.1594
Epoch 2/50
2636/2636 [==============================] - 39s 15ms/step - loss: 1.3654 - categorical_accuracy: 0.5945 - precision_5: 0.1290 - recall_5: 0.9491 - f1_score: 0.1813 - val_loss: 1.5518 - val_categorical_accuracy: 0.5455 - val_precision_5: 0.1230 - val_recall_5: 0.9326 - val_f1_score: 0.1675
Epoch 3/50
2636/2636 [==============================] - 40s 15ms/step - loss: 1.3211 - categorical_accuracy: 0.6177 - precision_5: 0.1284 - recall_5: 0.9554 - f1_score: 0.1901 - val_loss: 1.6260 - val_categorical_accuracy: 0.5621 - val_precision_5: 0.1239 - val_recall_5: 0.9286 - val_f1_score: 0.1765
Epoch 4/50
2636/2636 [==============================] - 37s 14ms/step - loss: 1.3152 - categorical_accuracy: 0.6109 - precision_5: 0.1283 - recall_5: 0.9572 - f1_score: 0.1971 - val_loss: 1.5842 - val_categorical_accuracy: 0.5486 - val_precision_5: 0.1235 - val_recall_5: 0.9306 - val_f1_score: 0.1736
Epoch 5/50
2636/2636 [==============================] - 40s 15ms/step - loss: 1.3028 - categorical_accuracy: 0.6187 - precision_5: 0.1281 - recall_5: 0.9593 - f1_score: 0.2018 - val_loss: 1.6329 - val_categorical_accuracy: 0.5534 - val_precision_5: 0.1256 - val_recall_5: 0.9193 - val_f1_score: 0.1783
Epoch 6/50
2636/2636 [==============================] - 36s 13ms/step - loss: 1.2920 - categorical_accuracy: 0.6262 - precision_5: 0.1300 - recall_5: 0.9584 - f1_score: 0.2049 - val_loss: 1.5742 - val_categorical_accuracy: 0.5747 - val_precision_5: 0.1215 - val_recall_5: 0.9303 - val_f1_score: 0.1723
Epoch 7/50
2636/2636 [==============================] - 41s 16ms/step - loss: 1.2820 - categorical_accuracy: 0.6260 - precision_5: 0.1305 - recall_5: 0.9562 - f1_score: 0.2090 - val_loss: 1.5755 - val_categorical_accuracy: 0.5829 - val_precision_5: 0.1264 - val_recall_5: 0.9332 - val_f1_score: 0.1857
Epoch 8/50
2636/2636 [==============================] - 40s 15ms/step - loss: 1.2745 - categorical_accuracy: 0.6346 - precision_5: 0.1316 - recall_5: 0.9565 - f1_score: 0.2145 - val_loss: 1.6308 - val_categorical_accuracy: 0.5762 - val_precision_5: 0.1261 - val_recall_5: 0.9228 - val_f1_score: 0.1772
Epoch 9/50
2636/2636 [==============================] - 36s 14ms/step - loss: 1.2715 - categorical_accuracy: 0.6337 - precision_5: 0.1322 - recall_5: 0.9559 - f1_score: 0.2159 - val_loss: 1.6406 - val_categorical_accuracy: 0.5852 - val_precision_5: 0.1262 - val_recall_5: 0.9209 - val_f1_score: 0.1785
Epoch 10/50
2636/2636 [==============================] - 47s 18ms/step - loss: 1.2650 - categorical_accuracy: 0.6351 - precision_5: 0.1339 - recall_5: 0.9543 - f1_score: 0.2197 - val_loss: 1.6375 - val_categorical_accuracy: 0.5703 - val_precision_5: 0.1267 - val_recall_5: 0.9258 - val_f1_score: 0.1820

 1 32 15 0.001 189 48 

Epoch 1/50
2643/2643 [==============================] - 39s 14ms/step - loss: 1.6313 - categorical_accuracy: 0.5340 - precision_6: 0.1335 - recall_6: 0.8912 - f1_score: 0.1693 - val_loss: 1.4402 - val_categorical_accuracy: 0.5769 - val_precision_6: 0.1439 - val_recall_6: 0.9371 - val_f1_score: 0.1882
Epoch 2/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.3920 - categorical_accuracy: 0.5822 - precision_6: 0.1356 - recall_6: 0.9475 - f1_score: 0.2016 - val_loss: 1.4028 - val_categorical_accuracy: 0.5565 - val_precision_6: 0.1375 - val_recall_6: 0.9493 - val_f1_score: 0.2137
Epoch 3/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.3664 - categorical_accuracy: 0.5921 - precision_6: 0.1339 - recall_6: 0.9496 - f1_score: 0.2227 - val_loss: 1.3373 - val_categorical_accuracy: 0.6184 - val_precision_6: 0.1382 - val_recall_6: 0.9648 - val_f1_score: 0.2118
Epoch 4/50
2643/2643 [==============================] - 38s 15ms/step - loss: 1.3431 - categorical_accuracy: 0.6036 - precision_6: 0.1337 - recall_6: 0.9511 - f1_score: 0.2400 - val_loss: 1.3674 - val_categorical_accuracy: 0.6170 - val_precision_6: 0.1358 - val_recall_6: 0.9611 - val_f1_score: 0.2005
Epoch 5/50
2643/2643 [==============================] - 38s 14ms/step - loss: 1.3281 - categorical_accuracy: 0.6116 - precision_6: 0.1335 - recall_6: 0.9529 - f1_score: 0.2327 - val_loss: 1.3767 - val_categorical_accuracy: 0.5867 - val_precision_6: 0.1367 - val_recall_6: 0.9548 - val_f1_score: 0.2090

 2 32 15 0.001 190 47 

Epoch 1/50
2512/2512 [==============================] - 39s 15ms/step - loss: 1.5887 - categorical_accuracy: 0.5373 - precision_7: 0.1307 - recall_7: 0.8887 - f1_score: 0.1596 - val_loss: 2.1039 - val_categorical_accuracy: 0.4683 - val_precision_7: 0.1167 - val_recall_7: 0.8594 - val_f1_score: 0.1496
Epoch 2/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.3432 - categorical_accuracy: 0.5968 - precision_7: 0.1324 - recall_7: 0.9510 - f1_score: 0.1919 - val_loss: 2.1386 - val_categorical_accuracy: 0.4801 - val_precision_7: 0.1240 - val_recall_7: 0.8848 - val_f1_score: 0.1637
Epoch 3/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.3180 - categorical_accuracy: 0.6043 - precision_7: 0.1316 - recall_7: 0.9548 - f1_score: 0.2023 - val_loss: 2.0110 - val_categorical_accuracy: 0.4772 - val_precision_7: 0.1227 - val_recall_7: 0.9162 - val_f1_score: 0.1624
Epoch 4/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.3014 - categorical_accuracy: 0.6139 - precision_7: 0.1305 - recall_7: 0.9537 - f1_score: 0.2097 - val_loss: 2.1472 - val_categorical_accuracy: 0.4650 - val_precision_7: 0.1259 - val_recall_7: 0.9185 - val_f1_score: 0.1640
Epoch 5/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2976 - categorical_accuracy: 0.6114 - precision_7: 0.1315 - recall_7: 0.9616 - f1_score: 0.2218 - val_loss: 2.2125 - val_categorical_accuracy: 0.4840 - val_precision_7: 0.1240 - val_recall_7: 0.8964 - val_f1_score: 0.1653
Epoch 6/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2838 - categorical_accuracy: 0.6262 - precision_7: 0.1323 - recall_7: 0.9542 - f1_score: 0.2187 - val_loss: 2.2631 - val_categorical_accuracy: 0.4479 - val_precision_7: 0.1253 - val_recall_7: 0.9262 - val_f1_score: 0.1640
Epoch 7/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2738 - categorical_accuracy: 0.6264 - precision_7: 0.1329 - recall_7: 0.9557 - f1_score: 0.2242 - val_loss: 2.1963 - val_categorical_accuracy: 0.4650 - val_precision_7: 0.1227 - val_recall_7: 0.9334 - val_f1_score: 0.1579
Epoch 8/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2742 - categorical_accuracy: 0.6246 - precision_7: 0.1327 - recall_7: 0.9554 - f1_score: 0.2293 - val_loss: 2.4059 - val_categorical_accuracy: 0.4664 - val_precision_7: 0.1278 - val_recall_7: 0.8853 - val_f1_score: 0.1684
Epoch 9/50
2512/2512 [==============================] - 39s 15ms/step - loss: 1.2609 - categorical_accuracy: 0.6354 - precision_7: 0.1364 - recall_7: 0.9576 - f1_score: 0.2281 - val_loss: 2.2724 - val_categorical_accuracy: 0.4759 - val_precision_7: 0.1226 - val_recall_7: 0.8977 - val_f1_score: 0.1687
Epoch 10/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2659 - categorical_accuracy: 0.6335 - precision_7: 0.1351 - recall_7: 0.9541 - f1_score: 0.2368 - val_loss: 2.3369 - val_categorical_accuracy: 0.4662 - val_precision_7: 0.1261 - val_recall_7: 0.9206 - val_f1_score: 0.1678
Epoch 11/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2613 - categorical_accuracy: 0.6197 - precision_7: 0.1360 - recall_7: 0.9534 - f1_score: 0.2392 - val_loss: 2.3875 - val_categorical_accuracy: 0.4774 - val_precision_7: 0.1268 - val_recall_7: 0.9026 - val_f1_score: 0.1672
Epoch 12/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2585 - categorical_accuracy: 0.6349 - precision_7: 0.1373 - recall_7: 0.9466 - f1_score: 0.2324 - val_loss: 2.3412 - val_categorical_accuracy: 0.4795 - val_precision_7: 0.1310 - val_recall_7: 0.9275 - val_f1_score: 0.1720
Epoch 13/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2517 - categorical_accuracy: 0.6364 - precision_7: 0.1381 - recall_7: 0.9481 - f1_score: 0.2405 - val_loss: 2.3580 - val_categorical_accuracy: 0.4568 - val_precision_7: 0.1239 - val_recall_7: 0.8737 - val_f1_score: 0.1670
Epoch 14/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2578 - categorical_accuracy: 0.6328 - precision_7: 0.1385 - recall_7: 0.9484 - f1_score: 0.2423 - val_loss: 2.4306 - val_categorical_accuracy: 0.4771 - val_precision_7: 0.1264 - val_recall_7: 0.8778 - val_f1_score: 0.1685
Epoch 15/50
2512/2512 [==============================] - 38s 15ms/step - loss: 1.2458 - categorical_accuracy: 0.6374 - precision_7: 0.1401 - recall_7: 0.9481 - f1_score: 0.2442 - val_loss: 2.4108 - val_categorical_accuracy: 0.4762 - val_precision_7: 0.1251 - val_recall_7: 0.8668 - val_f1_score: 0.1689

 3 32 15 0.001 190 47 

Epoch 1/50
2678/2678 [==============================] - 39s 14ms/step - loss: 1.6369 - categorical_accuracy: 0.5218 - precision_8: 0.1288 - recall_8: 0.8985 - f1_score: 0.1658 - val_loss: 1.5153 - val_categorical_accuracy: 0.5301 - val_precision_8: 0.1305 - val_recall_8: 0.9404 - val_f1_score: 0.1878
Epoch 2/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3784 - categorical_accuracy: 0.5848 - precision_8: 0.1323 - recall_8: 0.9605 - f1_score: 0.2017 - val_loss: 1.4893 - val_categorical_accuracy: 0.5445 - val_precision_8: 0.1332 - val_recall_8: 0.9482 - val_f1_score: 0.1977
Epoch 3/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3507 - categorical_accuracy: 0.5973 - precision_8: 0.1320 - recall_8: 0.9603 - f1_score: 0.2099 - val_loss: 1.4097 - val_categorical_accuracy: 0.6045 - val_precision_8: 0.1304 - val_recall_8: 0.9398 - val_f1_score: 0.1932
Epoch 4/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3271 - categorical_accuracy: 0.5998 - precision_8: 0.1334 - recall_8: 0.9651 - f1_score: 0.2253 - val_loss: 1.4649 - val_categorical_accuracy: 0.5762 - val_precision_8: 0.1298 - val_recall_8: 0.9526 - val_f1_score: 0.1955
Epoch 5/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3228 - categorical_accuracy: 0.6064 - precision_8: 0.1327 - recall_8: 0.9642 - f1_score: 0.2246 - val_loss: 1.3844 - val_categorical_accuracy: 0.6128 - val_precision_8: 0.1353 - val_recall_8: 0.9518 - val_f1_score: 0.2019
Epoch 6/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3115 - categorical_accuracy: 0.6123 - precision_8: 0.1341 - recall_8: 0.9660 - f1_score: 0.2345 - val_loss: 1.3734 - val_categorical_accuracy: 0.6127 - val_precision_8: 0.1314 - val_recall_8: 0.9455 - val_f1_score: 0.1959
Epoch 7/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.3080 - categorical_accuracy: 0.6122 - precision_8: 0.1356 - recall_8: 0.9639 - f1_score: 0.2428 - val_loss: 1.3344 - val_categorical_accuracy: 0.6367 - val_precision_8: 0.1356 - val_recall_8: 0.9444 - val_f1_score: 0.2065
Epoch 8/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.2930 - categorical_accuracy: 0.6183 - precision_8: 0.1358 - recall_8: 0.9613 - f1_score: 0.2466 - val_loss: 1.3676 - val_categorical_accuracy: 0.5957 - val_precision_8: 0.1354 - val_recall_8: 0.9425 - val_f1_score: 0.2057
Epoch 9/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.2968 - categorical_accuracy: 0.6173 - precision_8: 0.1373 - recall_8: 0.9599 - f1_score: 0.2601 - val_loss: 1.4133 - val_categorical_accuracy: 0.5779 - val_precision_8: 0.1341 - val_recall_8: 0.9377 - val_f1_score: 0.2073
Epoch 10/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.2909 - categorical_accuracy: 0.6170 - precision_8: 0.1379 - recall_8: 0.9613 - f1_score: 0.2594 - val_loss: 1.3471 - val_categorical_accuracy: 0.6150 - val_precision_8: 0.1371 - val_recall_8: 0.9385 - val_f1_score: 0.2018
Epoch 11/50
2678/2678 [==============================] - 40s 15ms/step - loss: 1.2856 - categorical_accuracy: 0.6215 - precision_8: 0.1403 - recall_8: 0.9603 - f1_score: 0.2675 - val_loss: 1.3609 - val_categorical_accuracy: 0.6345 - val_precision_8: 0.1401 - val_recall_8: 0.9442 - val_f1_score: 0.2157
Epoch 12/50
2678/2678 [==============================] - 39s 15ms/step - loss: 1.2865 - categorical_accuracy: 0.6189 - precision_8: 0.1390 - recall_8: 0.9613 - f1_score: 0.2713 - val_loss: 1.3613 - val_categorical_accuracy: 0.6000 - val_precision_8: 0.1417 - val_recall_8: 0.9320 - val_f1_score: 0.2214
Epoch 13/50
2678/2678 [==============================] - 37s 14ms/step - loss: 1.2840 - categorical_accuracy: 0.6287 - precision_8: 0.1411 - recall_8: 0.9519 - f1_score: 0.2638 - val_loss: 1.3594 - val_categorical_accuracy: 0.6226 - val_precision_8: 0.1387 - val_recall_8: 0.9323 - val_f1_score: 0.2085
Epoch 14/50
2678/2678 [==============================] - 39s 15ms/step - loss: 1.2796 - categorical_accuracy: 0.6241 - precision_8: 0.1405 - recall_8: 0.9524 - f1_score: 0.2741 - val_loss: 1.3435 - val_categorical_accuracy: 0.6356 - val_precision_8: 0.1415 - val_recall_8: 0.9317 - val_f1_score: 0.2194
Epoch 15/50
2678/2678 [==============================] - 38s 14ms/step - loss: 1.2792 - categorical_accuracy: 0.6251 - precision_8: 0.1428 - recall_8: 0.9533 - f1_score: 0.2816 - val_loss: 1.3648 - val_categorical_accuracy: 0.6264 - val_precision_8: 0.1411 - val_recall_8: 0.9301 - val_f1_score: 0.2145

 4 32 15 0.001 190 47 

Epoch 1/50
2679/2679 [==============================] - 39s 14ms/step - loss: 1.6523 - categorical_accuracy: 0.5177 - precision_9: 0.1230 - recall_9: 0.8855 - f1_score: 0.1589 - val_loss: 1.3559 - val_categorical_accuracy: 0.5835 - val_precision_9: 0.1261 - val_recall_9: 0.9467 - val_f1_score: 0.1672
Epoch 2/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3897 - categorical_accuracy: 0.5780 - precision_9: 0.1292 - recall_9: 0.9470 - f1_score: 0.1887 - val_loss: 1.4041 - val_categorical_accuracy: 0.5973 - val_precision_9: 0.1297 - val_recall_9: 0.9551 - val_f1_score: 0.1792
Epoch 3/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3788 - categorical_accuracy: 0.5882 - precision_9: 0.1289 - recall_9: 0.9531 - f1_score: 0.2030 - val_loss: 1.3189 - val_categorical_accuracy: 0.6165 - val_precision_9: 0.1304 - val_recall_9: 0.9494 - val_f1_score: 0.1762
Epoch 4/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3455 - categorical_accuracy: 0.6002 - precision_9: 0.1305 - recall_9: 0.9538 - f1_score: 0.2175 - val_loss: 1.3091 - val_categorical_accuracy: 0.6245 - val_precision_9: 0.1280 - val_recall_9: 0.9459 - val_f1_score: 0.1813
Epoch 5/50
2679/2679 [==============================] - 39s 15ms/step - loss: 1.3368 - categorical_accuracy: 0.6000 - precision_9: 0.1316 - recall_9: 0.9545 - f1_score: 0.2241 - val_loss: 1.3298 - val_categorical_accuracy: 0.6189 - val_precision_9: 0.1347 - val_recall_9: 0.9419 - val_f1_score: 0.1821
Epoch 6/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3366 - categorical_accuracy: 0.6065 - precision_9: 0.1330 - recall_9: 0.9506 - f1_score: 0.2287 - val_loss: 1.2623 - val_categorical_accuracy: 0.6596 - val_precision_9: 0.1313 - val_recall_9: 0.9540 - val_f1_score: 0.1845
Epoch 7/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3161 - categorical_accuracy: 0.6190 - precision_9: 0.1325 - recall_9: 0.9519 - f1_score: 0.2391 - val_loss: 1.3004 - val_categorical_accuracy: 0.6308 - val_precision_9: 0.1359 - val_recall_9: 0.9515 - val_f1_score: 0.1928
Epoch 8/50
2679/2679 [==============================] - 39s 14ms/step - loss: 1.3123 - categorical_accuracy: 0.6140 - precision_9: 0.1349 - recall_9: 0.9531 - f1_score: 0.2408 - val_loss: 1.3326 - val_categorical_accuracy: 0.6348 - val_precision_9: 0.1346 - val_recall_9: 0.9488 - val_f1_score: 0.2008
Epoch 9/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3177 - categorical_accuracy: 0.6115 - precision_9: 0.1346 - recall_9: 0.9505 - f1_score: 0.2495 - val_loss: 1.3136 - val_categorical_accuracy: 0.6320 - val_precision_9: 0.1380 - val_recall_9: 0.9475 - val_f1_score: 0.1936
Epoch 10/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3051 - categorical_accuracy: 0.6152 - precision_9: 0.1372 - recall_9: 0.9512 - f1_score: 0.2584 - val_loss: 1.2731 - val_categorical_accuracy: 0.6247 - val_precision_9: 0.1376 - val_recall_9: 0.9524 - val_f1_score: 0.2089
Epoch 11/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.3026 - categorical_accuracy: 0.6143 - precision_9: 0.1372 - recall_9: 0.9459 - f1_score: 0.2607 - val_loss: 1.2847 - val_categorical_accuracy: 0.6625 - val_precision_9: 0.1389 - val_recall_9: 0.9447 - val_f1_score: 0.2080
Epoch 12/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.2917 - categorical_accuracy: 0.6248 - precision_9: 0.1380 - recall_9: 0.9444 - f1_score: 0.2482 - val_loss: 1.2855 - val_categorical_accuracy: 0.6687 - val_precision_9: 0.1403 - val_recall_9: 0.9480 - val_f1_score: 0.2107
Epoch 13/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.2995 - categorical_accuracy: 0.6208 - precision_9: 0.1402 - recall_9: 0.9451 - f1_score: 0.2697 - val_loss: 1.2269 - val_categorical_accuracy: 0.6753 - val_precision_9: 0.1381 - val_recall_9: 0.9464 - val_f1_score: 0.2043
Epoch 14/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.2921 - categorical_accuracy: 0.6227 - precision_9: 0.1419 - recall_9: 0.9458 - f1_score: 0.2767 - val_loss: 1.2962 - val_categorical_accuracy: 0.6529 - val_precision_9: 0.1431 - val_recall_9: 0.9460 - val_f1_score: 0.2028
Epoch 15/50
2679/2679 [==============================] - 38s 14ms/step - loss: 1.2925 - categorical_accuracy: 0.6263 - precision_9: 0.1416 - recall_9: 0.9429 - f1_score: 0.2728 - val_loss: 1.2496 - val_categorical_accuracy: 0.6617 - val_precision_9: 0.1426 - val_recall_9: 0.9313 - val_f1_score: 0.2070

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

 0 32 30 0.0001 189 48 

Epoch 1/50
2584/2584 [==============================] - 72s 27ms/step - loss: 2.1609 - categorical_accuracy: 0.3926 - precision_10: 0.1043 - recall_10: 0.7741 - f1_score: 0.1456 - val_loss: 1.9377 - val_categorical_accuracy: 0.4388 - val_precision_10: 0.1139 - val_recall_10: 0.8315 - val_f1_score: 0.1411
Epoch 2/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.5966 - categorical_accuracy: 0.5307 - precision_10: 0.1270 - recall_10: 0.8803 - f1_score: 0.1594 - val_loss: 1.7016 - val_categorical_accuracy: 0.4890 - val_precision_10: 0.1276 - val_recall_10: 0.8881 - val_f1_score: 0.1533
Epoch 3/50
2584/2584 [==============================] - 43s 16ms/step - loss: 1.4295 - categorical_accuracy: 0.5676 - precision_10: 0.1335 - recall_10: 0.9166 - f1_score: 0.1682 - val_loss: 1.6125 - val_categorical_accuracy: 0.5074 - val_precision_10: 0.1290 - val_recall_10: 0.9097 - val_f1_score: 0.1595
Epoch 4/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.3568 - categorical_accuracy: 0.5899 - precision_10: 0.1331 - recall_10: 0.9338 - f1_score: 0.1749 - val_loss: 1.5847 - val_categorical_accuracy: 0.5190 - val_precision_10: 0.1305 - val_recall_10: 0.9202 - val_f1_score: 0.1691
Epoch 5/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.3173 - categorical_accuracy: 0.5911 - precision_10: 0.1334 - recall_10: 0.9477 - f1_score: 0.1817 - val_loss: 1.5556 - val_categorical_accuracy: 0.5290 - val_precision_10: 0.1293 - val_recall_10: 0.9240 - val_f1_score: 0.1692
Epoch 6/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2902 - categorical_accuracy: 0.6060 - precision_10: 0.1330 - recall_10: 0.9510 - f1_score: 0.1882 - val_loss: 1.5486 - val_categorical_accuracy: 0.5268 - val_precision_10: 0.1300 - val_recall_10: 0.9294 - val_f1_score: 0.1764
Epoch 7/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2708 - categorical_accuracy: 0.6081 - precision_10: 0.1340 - recall_10: 0.9567 - f1_score: 0.1898 - val_loss: 1.5340 - val_categorical_accuracy: 0.5377 - val_precision_10: 0.1289 - val_recall_10: 0.9301 - val_f1_score: 0.1827
Epoch 8/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2560 - categorical_accuracy: 0.6107 - precision_10: 0.1335 - recall_10: 0.9598 - f1_score: 0.1917 - val_loss: 1.5375 - val_categorical_accuracy: 0.5349 - val_precision_10: 0.1281 - val_recall_10: 0.9341 - val_f1_score: 0.1859
Epoch 9/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2446 - categorical_accuracy: 0.6178 - precision_10: 0.1334 - recall_10: 0.9600 - f1_score: 0.1926 - val_loss: 1.5376 - val_categorical_accuracy: 0.5446 - val_precision_10: 0.1292 - val_recall_10: 0.9380 - val_f1_score: 0.1907
Epoch 10/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2315 - categorical_accuracy: 0.6250 - precision_10: 0.1321 - recall_10: 0.9623 - f1_score: 0.1947 - val_loss: 1.5454 - val_categorical_accuracy: 0.5467 - val_precision_10: 0.1291 - val_recall_10: 0.9397 - val_f1_score: 0.1896
Epoch 11/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2254 - categorical_accuracy: 0.6202 - precision_10: 0.1325 - recall_10: 0.9643 - f1_score: 0.1958 - val_loss: 1.5498 - val_categorical_accuracy: 0.5483 - val_precision_10: 0.1285 - val_recall_10: 0.9387 - val_f1_score: 0.1903
Epoch 12/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2198 - categorical_accuracy: 0.6263 - precision_10: 0.1325 - recall_10: 0.9638 - f1_score: 0.1973 - val_loss: 1.5508 - val_categorical_accuracy: 0.5458 - val_precision_10: 0.1274 - val_recall_10: 0.9399 - val_f1_score: 0.1903

 1 32 30 0.0001 189 48 

Epoch 1/50
2595/2595 [==============================] - 43s 16ms/step - loss: 2.1147 - categorical_accuracy: 0.4145 - precision_11: 0.1118 - recall_11: 0.7779 - f1_score: 0.1538 - val_loss: 1.7829 - val_categorical_accuracy: 0.5070 - val_precision_11: 0.1228 - val_recall_11: 0.8431 - val_f1_score: 0.1587
Epoch 2/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.5859 - categorical_accuracy: 0.5459 - precision_11: 0.1328 - recall_11: 0.8937 - f1_score: 0.1679 - val_loss: 1.5470 - val_categorical_accuracy: 0.5488 - val_precision_11: 0.1399 - val_recall_11: 0.8955 - val_f1_score: 0.1715
Epoch 3/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.4412 - categorical_accuracy: 0.5747 - precision_11: 0.1396 - recall_11: 0.9205 - f1_score: 0.1780 - val_loss: 1.4706 - val_categorical_accuracy: 0.5704 - val_precision_11: 0.1370 - val_recall_11: 0.9151 - val_f1_score: 0.1735
Epoch 4/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3740 - categorical_accuracy: 0.5911 - precision_11: 0.1385 - recall_11: 0.9322 - f1_score: 0.1839 - val_loss: 1.4256 - val_categorical_accuracy: 0.5617 - val_precision_11: 0.1367 - val_recall_11: 0.9251 - val_f1_score: 0.1825
Epoch 5/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3337 - categorical_accuracy: 0.5953 - precision_11: 0.1382 - recall_11: 0.9391 - f1_score: 0.1913 - val_loss: 1.3877 - val_categorical_accuracy: 0.5698 - val_precision_11: 0.1382 - val_recall_11: 0.9344 - val_f1_score: 0.1875
Epoch 6/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3087 - categorical_accuracy: 0.5995 - precision_11: 0.1369 - recall_11: 0.9426 - f1_score: 0.1979 - val_loss: 1.3644 - val_categorical_accuracy: 0.5788 - val_precision_11: 0.1379 - val_recall_11: 0.9386 - val_f1_score: 0.1903
Epoch 7/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2890 - categorical_accuracy: 0.6025 - precision_11: 0.1370 - recall_11: 0.9448 - f1_score: 0.2082 - val_loss: 1.3620 - val_categorical_accuracy: 0.5711 - val_precision_11: 0.1366 - val_recall_11: 0.9461 - val_f1_score: 0.1918
Epoch 8/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2758 - categorical_accuracy: 0.6089 - precision_11: 0.1361 - recall_11: 0.9488 - f1_score: 0.2093 - val_loss: 1.3273 - val_categorical_accuracy: 0.5816 - val_precision_11: 0.1360 - val_recall_11: 0.9475 - val_f1_score: 0.1929
Epoch 9/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2635 - categorical_accuracy: 0.6105 - precision_11: 0.1354 - recall_11: 0.9500 - f1_score: 0.2154 - val_loss: 1.3259 - val_categorical_accuracy: 0.5858 - val_precision_11: 0.1363 - val_recall_11: 0.9510 - val_f1_score: 0.1953
Epoch 10/50
2595/2595 [==============================] - 42s 16ms/step - loss: 1.2527 - categorical_accuracy: 0.6169 - precision_11: 0.1355 - recall_11: 0.9511 - f1_score: 0.2210 - val_loss: 1.3275 - val_categorical_accuracy: 0.5934 - val_precision_11: 0.1352 - val_recall_11: 0.9501 - val_f1_score: 0.1948
Epoch 11/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2469 - categorical_accuracy: 0.6157 - precision_11: 0.1351 - recall_11: 0.9519 - f1_score: 0.2266 - val_loss: 1.3152 - val_categorical_accuracy: 0.5918 - val_precision_11: 0.1365 - val_recall_11: 0.9477 - val_f1_score: 0.1976
Epoch 12/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2399 - categorical_accuracy: 0.6205 - precision_11: 0.1354 - recall_11: 0.9530 - f1_score: 0.2318 - val_loss: 1.3173 - val_categorical_accuracy: 0.5863 - val_precision_11: 0.1364 - val_recall_11: 0.9531 - val_f1_score: 0.1977
Epoch 13/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2329 - categorical_accuracy: 0.6272 - precision_11: 0.1343 - recall_11: 0.9528 - f1_score: 0.2309 - val_loss: 1.3119 - val_categorical_accuracy: 0.5788 - val_precision_11: 0.1365 - val_recall_11: 0.9557 - val_f1_score: 0.2007
Epoch 14/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2282 - categorical_accuracy: 0.6210 - precision_11: 0.1341 - recall_11: 0.9547 - f1_score: 0.2361 - val_loss: 1.3061 - val_categorical_accuracy: 0.5908 - val_precision_11: 0.1360 - val_recall_11: 0.9581 - val_f1_score: 0.1978
Epoch 15/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2256 - categorical_accuracy: 0.6237 - precision_11: 0.1353 - recall_11: 0.9562 - f1_score: 0.2411 - val_loss: 1.3034 - val_categorical_accuracy: 0.5965 - val_precision_11: 0.1348 - val_recall_11: 0.9546 - val_f1_score: 0.1973
Epoch 16/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2173 - categorical_accuracy: 0.6312 - precision_11: 0.1345 - recall_11: 0.9553 - f1_score: 0.2425 - val_loss: 1.3094 - val_categorical_accuracy: 0.5970 - val_precision_11: 0.1350 - val_recall_11: 0.9555 - val_f1_score: 0.2012
Epoch 17/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2180 - categorical_accuracy: 0.6295 - precision_11: 0.1338 - recall_11: 0.9557 - f1_score: 0.2392 - val_loss: 1.2952 - val_categorical_accuracy: 0.6124 - val_precision_11: 0.1369 - val_recall_11: 0.9618 - val_f1_score: 0.2050
Epoch 18/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2146 - categorical_accuracy: 0.6279 - precision_11: 0.1341 - recall_11: 0.9556 - f1_score: 0.2489 - val_loss: 1.2954 - val_categorical_accuracy: 0.6013 - val_precision_11: 0.1371 - val_recall_11: 0.9631 - val_f1_score: 0.2036
Epoch 19/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2108 - categorical_accuracy: 0.6329 - precision_11: 0.1340 - recall_11: 0.9583 - f1_score: 0.2476 - val_loss: 1.3029 - val_categorical_accuracy: 0.6026 - val_precision_11: 0.1362 - val_recall_11: 0.9605 - val_f1_score: 0.2062
Epoch 20/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2114 - categorical_accuracy: 0.6330 - precision_11: 0.1342 - recall_11: 0.9559 - f1_score: 0.2505 - val_loss: 1.2968 - val_categorical_accuracy: 0.6000 - val_precision_11: 0.1343 - val_recall_11: 0.9582 - val_f1_score: 0.2014
Epoch 21/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2064 - categorical_accuracy: 0.6323 - precision_11: 0.1334 - recall_11: 0.9562 - f1_score: 0.2491 - val_loss: 1.2942 - val_categorical_accuracy: 0.6244 - val_precision_11: 0.1365 - val_recall_11: 0.9593 - val_f1_score: 0.2035
Epoch 22/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2062 - categorical_accuracy: 0.6324 - precision_11: 0.1343 - recall_11: 0.9573 - f1_score: 0.2556 - val_loss: 1.2832 - val_categorical_accuracy: 0.6149 - val_precision_11: 0.1377 - val_recall_11: 0.9616 - val_f1_score: 0.2101
Epoch 23/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2019 - categorical_accuracy: 0.6365 - precision_11: 0.1344 - recall_11: 0.9579 - f1_score: 0.2566 - val_loss: 1.2974 - val_categorical_accuracy: 0.6024 - val_precision_11: 0.1374 - val_recall_11: 0.9594 - val_f1_score: 0.2116
Epoch 24/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2013 - categorical_accuracy: 0.6354 - precision_11: 0.1339 - recall_11: 0.9573 - f1_score: 0.2592 - val_loss: 1.2866 - val_categorical_accuracy: 0.6296 - val_precision_11: 0.1350 - val_recall_11: 0.9584 - val_f1_score: 0.2032
Epoch 25/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.2011 - categorical_accuracy: 0.6382 - precision_11: 0.1336 - recall_11: 0.9578 - f1_score: 0.2569 - val_loss: 1.2804 - val_categorical_accuracy: 0.6163 - val_precision_11: 0.1353 - val_recall_11: 0.9584 - val_f1_score: 0.2031
Epoch 26/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.1988 - categorical_accuracy: 0.6409 - precision_11: 0.1344 - recall_11: 0.9584 - f1_score: 0.2556 - val_loss: 1.2815 - val_categorical_accuracy: 0.6188 - val_precision_11: 0.1356 - val_recall_11: 0.9635 - val_f1_score: 0.2041

 2 32 30 0.0001 190 47 

Epoch 1/50
2461/2461 [==============================] - 42s 17ms/step - loss: 2.1222 - categorical_accuracy: 0.4194 - precision_12: 0.0998 - recall_12: 0.7756 - f1_score: 0.1348 - val_loss: 2.0653 - val_categorical_accuracy: 0.4129 - val_precision_12: 0.1002 - val_recall_12: 0.7829 - val_f1_score: 0.1385
Epoch 2/50
2461/2461 [==============================] - 41s 16ms/step - loss: 1.5725 - categorical_accuracy: 0.5421 - precision_12: 0.1228 - recall_12: 0.8850 - f1_score: 0.1515 - val_loss: 2.0214 - val_categorical_accuracy: 0.4397 - val_precision_12: 0.1127 - val_recall_12: 0.8205 - val_f1_score: 0.1415
Epoch 3/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.4121 - categorical_accuracy: 0.5745 - precision_12: 0.1291 - recall_12: 0.9177 - f1_score: 0.1597 - val_loss: 2.0648 - val_categorical_accuracy: 0.4464 - val_precision_12: 0.1164 - val_recall_12: 0.8403 - val_f1_score: 0.1445
Epoch 4/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.3438 - categorical_accuracy: 0.5871 - precision_12: 0.1315 - recall_12: 0.9338 - f1_score: 0.1672 - val_loss: 2.0563 - val_categorical_accuracy: 0.4436 - val_precision_12: 0.1177 - val_recall_12: 0.8530 - val_f1_score: 0.1462
Epoch 5/50
2461/2461 [==============================] - 41s 16ms/step - loss: 1.3044 - categorical_accuracy: 0.5904 - precision_12: 0.1322 - recall_12: 0.9443 - f1_score: 0.1732 - val_loss: 2.0924 - val_categorical_accuracy: 0.4509 - val_precision_12: 0.1169 - val_recall_12: 0.8624 - val_f1_score: 0.1461
Epoch 6/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2739 - categorical_accuracy: 0.6002 - precision_12: 0.1309 - recall_12: 0.9485 - f1_score: 0.1766 - val_loss: 2.1047 - val_categorical_accuracy: 0.4480 - val_precision_12: 0.1181 - val_recall_12: 0.8648 - val_f1_score: 0.1513
Epoch 7/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2568 - categorical_accuracy: 0.6037 - precision_12: 0.1312 - recall_12: 0.9525 - f1_score: 0.1815 - val_loss: 2.1686 - val_categorical_accuracy: 0.4560 - val_precision_12: 0.1177 - val_recall_12: 0.8793 - val_f1_score: 0.1530
Epoch 8/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2431 - categorical_accuracy: 0.6105 - precision_12: 0.1307 - recall_12: 0.9543 - f1_score: 0.1839 - val_loss: 2.1885 - val_categorical_accuracy: 0.4584 - val_precision_12: 0.1177 - val_recall_12: 0.8816 - val_f1_score: 0.1541
Epoch 9/50
2461/2461 [==============================] - 42s 17ms/step - loss: 1.2304 - categorical_accuracy: 0.6092 - precision_12: 0.1306 - recall_12: 0.9569 - f1_score: 0.1845 - val_loss: 2.2266 - val_categorical_accuracy: 0.4534 - val_precision_12: 0.1179 - val_recall_12: 0.8866 - val_f1_score: 0.1541
Epoch 10/50
2461/2461 [==============================] - 41s 16ms/step - loss: 1.2242 - categorical_accuracy: 0.6176 - precision_12: 0.1304 - recall_12: 0.9580 - f1_score: 0.1890 - val_loss: 2.2516 - val_categorical_accuracy: 0.4596 - val_precision_12: 0.1191 - val_recall_12: 0.8958 - val_f1_score: 0.1550
Epoch 11/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2166 - categorical_accuracy: 0.6189 - precision_12: 0.1302 - recall_12: 0.9595 - f1_score: 0.1912 - val_loss: 2.2669 - val_categorical_accuracy: 0.4572 - val_precision_12: 0.1189 - val_recall_12: 0.9031 - val_f1_score: 0.1556
Epoch 12/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2098 - categorical_accuracy: 0.6190 - precision_12: 0.1296 - recall_12: 0.9602 - f1_score: 0.1911 - val_loss: 2.2965 - val_categorical_accuracy: 0.4598 - val_precision_12: 0.1201 - val_recall_12: 0.9040 - val_f1_score: 0.1568
Epoch 13/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2058 - categorical_accuracy: 0.6232 - precision_12: 0.1295 - recall_12: 0.9609 - f1_score: 0.1950 - val_loss: 2.3365 - val_categorical_accuracy: 0.4664 - val_precision_12: 0.1194 - val_recall_12: 0.8989 - val_f1_score: 0.1565
Epoch 14/50
2461/2461 [==============================] - 41s 17ms/step - loss: 1.2000 - categorical_accuracy: 0.6250 - precision_12: 0.1294 - recall_12: 0.9623 - f1_score: 0.1966 - val_loss: 2.3597 - val_categorical_accuracy: 0.4596 - val_precision_12: 0.1172 - val_recall_12: 0.8818 - val_f1_score: 0.1547
Epoch 15/50
2461/2461 [==============================] - 40s 16ms/step - loss: 1.1934 - categorical_accuracy: 0.6285 - precision_12: 0.1299 - recall_12: 0.9615 - f1_score: 0.2002 - val_loss: 2.3849 - val_categorical_accuracy: 0.4615 - val_precision_12: 0.1164 - val_recall_12: 0.8818 - val_f1_score: 0.1538

 3 32 30 0.0001 190 47 

Epoch 1/50
2627/2627 [==============================] - 43s 16ms/step - loss: 2.1257 - categorical_accuracy: 0.4160 - precision_13: 0.1072 - recall_13: 0.7821 - f1_score: 0.1490 - val_loss: 1.7528 - val_categorical_accuracy: 0.5236 - val_precision_13: 0.1207 - val_recall_13: 0.8439 - val_f1_score: 0.1521
Epoch 2/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.6143 - categorical_accuracy: 0.5359 - precision_13: 0.1272 - recall_13: 0.8772 - f1_score: 0.1609 - val_loss: 1.5304 - val_categorical_accuracy: 0.5660 - val_precision_13: 0.1341 - val_recall_13: 0.8963 - val_f1_score: 0.1615
Epoch 3/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.4544 - categorical_accuracy: 0.5697 - precision_13: 0.1343 - recall_13: 0.9146 - f1_score: 0.1712 - val_loss: 1.4320 - val_categorical_accuracy: 0.5830 - val_precision_13: 0.1336 - val_recall_13: 0.9181 - val_f1_score: 0.1697
Epoch 4/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.3813 - categorical_accuracy: 0.5784 - precision_13: 0.1342 - recall_13: 0.9360 - f1_score: 0.1801 - val_loss: 1.4120 - val_categorical_accuracy: 0.5847 - val_precision_13: 0.1326 - val_recall_13: 0.9250 - val_f1_score: 0.1730
Epoch 5/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.3369 - categorical_accuracy: 0.5916 - precision_13: 0.1340 - recall_13: 0.9432 - f1_score: 0.1890 - val_loss: 1.3982 - val_categorical_accuracy: 0.5816 - val_precision_13: 0.1321 - val_recall_13: 0.9321 - val_f1_score: 0.1809
Epoch 6/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.3114 - categorical_accuracy: 0.5960 - precision_13: 0.1341 - recall_13: 0.9523 - f1_score: 0.1983 - val_loss: 1.3728 - val_categorical_accuracy: 0.5826 - val_precision_13: 0.1317 - val_recall_13: 0.9335 - val_f1_score: 0.1834
Epoch 7/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.2892 - categorical_accuracy: 0.6033 - precision_13: 0.1340 - recall_13: 0.9561 - f1_score: 0.2062 - val_loss: 1.3574 - val_categorical_accuracy: 0.5975 - val_precision_13: 0.1316 - val_recall_13: 0.9356 - val_f1_score: 0.1844
Epoch 8/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.2745 - categorical_accuracy: 0.6054 - precision_13: 0.1344 - recall_13: 0.9591 - f1_score: 0.2104 - val_loss: 1.3413 - val_categorical_accuracy: 0.5987 - val_precision_13: 0.1315 - val_recall_13: 0.9379 - val_f1_score: 0.1887
Epoch 9/50
2627/2627 [==============================] - 42s 16ms/step - loss: 1.2618 - categorical_accuracy: 0.6083 - precision_13: 0.1333 - recall_13: 0.9623 - f1_score: 0.2186 - val_loss: 1.3354 - val_categorical_accuracy: 0.6062 - val_precision_13: 0.1327 - val_recall_13: 0.9381 - val_f1_score: 0.1914
Epoch 10/50
2627/2627 [==============================] - 43s 16ms/step - loss: 1.2476 - categorical_accuracy: 0.6172 - precision_13: 0.1342 - recall_13: 0.9635 - f1_score: 0.2181 - val_loss: 1.3316 - val_categorical_accuracy: 0.5989 - val_precision_13: 0.1329 - val_recall_13: 0.9403 - val_f1_score: 0.1923
Epoch 11/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.2437 - categorical_accuracy: 0.6155 - precision_13: 0.1348 - recall_13: 0.9660 - f1_score: 0.2239 - val_loss: 1.3380 - val_categorical_accuracy: 0.5907 - val_precision_13: 0.1329 - val_recall_13: 0.9429 - val_f1_score: 0.1975
Epoch 12/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.2348 - categorical_accuracy: 0.6175 - precision_13: 0.1337 - recall_13: 0.9679 - f1_score: 0.2251 - val_loss: 1.3131 - val_categorical_accuracy: 0.6111 - val_precision_13: 0.1322 - val_recall_13: 0.9426 - val_f1_score: 0.1931
Epoch 13/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.2309 - categorical_accuracy: 0.6201 - precision_13: 0.1339 - recall_13: 0.9685 - f1_score: 0.2293 - val_loss: 1.3360 - val_categorical_accuracy: 0.5978 - val_precision_13: 0.1322 - val_recall_13: 0.9421 - val_f1_score: 0.1960
Epoch 14/50
2627/2627 [==============================] - 41s 16ms/step - loss: 1.2267 - categorical_accuracy: 0.6227 - precision_13: 0.1342 - recall_13: 0.9692 - f1_score: 0.2340 - val_loss: 1.3160 - val_categorical_accuracy: 0.6037 - val_precision_13: 0.1319 - val_recall_13: 0.9426 - val_f1_score: 0.1947

 4 32 30 0.0001 190 47 

Epoch 1/50
2625/2625 [==============================] - 42s 16ms/step - loss: 2.1486 - categorical_accuracy: 0.4036 - precision_14: 0.1061 - recall_14: 0.7775 - f1_score: 0.1485 - val_loss: 1.6783 - val_categorical_accuracy: 0.5306 - val_precision_14: 0.1234 - val_recall_14: 0.8706 - val_f1_score: 0.1536
Epoch 2/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.6196 - categorical_accuracy: 0.5343 - precision_14: 0.1270 - recall_14: 0.8828 - f1_score: 0.1626 - val_loss: 1.4649 - val_categorical_accuracy: 0.5741 - val_precision_14: 0.1373 - val_recall_14: 0.9140 - val_f1_score: 0.1613
Epoch 3/50
2625/2625 [==============================] - 41s 16ms/step - loss: 1.4577 - categorical_accuracy: 0.5630 - precision_14: 0.1355 - recall_14: 0.9192 - f1_score: 0.1732 - val_loss: 1.3848 - val_categorical_accuracy: 0.5892 - val_precision_14: 0.1368 - val_recall_14: 0.9295 - val_f1_score: 0.1684
Epoch 4/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.3877 - categorical_accuracy: 0.5770 - precision_14: 0.1352 - recall_14: 0.9335 - f1_score: 0.1836 - val_loss: 1.3539 - val_categorical_accuracy: 0.5852 - val_precision_14: 0.1367 - val_recall_14: 0.9382 - val_f1_score: 0.1757
Epoch 5/50
2625/2625 [==============================] - 41s 15ms/step - loss: 1.3467 - categorical_accuracy: 0.5895 - precision_14: 0.1342 - recall_14: 0.9376 - f1_score: 0.1865 - val_loss: 1.3210 - val_categorical_accuracy: 0.6000 - val_precision_14: 0.1356 - val_recall_14: 0.9439 - val_f1_score: 0.1799
Epoch 6/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.3201 - categorical_accuracy: 0.5900 - precision_14: 0.1332 - recall_14: 0.9444 - f1_score: 0.1904 - val_loss: 1.3066 - val_categorical_accuracy: 0.5991 - val_precision_14: 0.1358 - val_recall_14: 0.9472 - val_f1_score: 0.1790
Epoch 7/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.3021 - categorical_accuracy: 0.5912 - precision_14: 0.1340 - recall_14: 0.9506 - f1_score: 0.2058 - val_loss: 1.2839 - val_categorical_accuracy: 0.6197 - val_precision_14: 0.1334 - val_recall_14: 0.9478 - val_f1_score: 0.1787
Epoch 8/50
2625/2625 [==============================] - 41s 15ms/step - loss: 1.2864 - categorical_accuracy: 0.5987 - precision_14: 0.1336 - recall_14: 0.9528 - f1_score: 0.2085 - val_loss: 1.2799 - val_categorical_accuracy: 0.6105 - val_precision_14: 0.1352 - val_recall_14: 0.9512 - val_f1_score: 0.1846
Epoch 9/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.2754 - categorical_accuracy: 0.6012 - precision_14: 0.1329 - recall_14: 0.9523 - f1_score: 0.2058 - val_loss: 1.2717 - val_categorical_accuracy: 0.6185 - val_precision_14: 0.1339 - val_recall_14: 0.9539 - val_f1_score: 0.1826
Epoch 10/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.2667 - categorical_accuracy: 0.6091 - precision_14: 0.1320 - recall_14: 0.9553 - f1_score: 0.2141 - val_loss: 1.2497 - val_categorical_accuracy: 0.6152 - val_precision_14: 0.1338 - val_recall_14: 0.9543 - val_f1_score: 0.1809
Epoch 11/50
2625/2625 [==============================] - 40s 15ms/step - loss: 1.2572 - categorical_accuracy: 0.6083 - precision_14: 0.1324 - recall_14: 0.9567 - f1_score: 0.2109 - val_loss: 1.2476 - val_categorical_accuracy: 0.6246 - val_precision_14: 0.1318 - val_recall_14: 0.9539 - val_f1_score: 0.1802

 0 32 30 0.001 189 48 

Epoch 1/50

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

2584/2584 [==============================] - 42s 16ms/step - loss: 1.6169 - categorical_accuracy: 0.5406 - precision_15: 0.1329 - recall_15: 0.8979 - f1_score: 0.1701 - val_loss: 1.5946 - val_categorical_accuracy: 0.5549 - val_precision_15: 0.1326 - val_recall_15: 0.9290 - val_f1_score: 0.1893
Epoch 2/50
2584/2584 [==============================] - 40s 16ms/step - loss: 1.3807 - categorical_accuracy: 0.5882 - precision_15: 0.1363 - recall_15: 0.9530 - f1_score: 0.1983 - val_loss: 1.5753 - val_categorical_accuracy: 0.5482 - val_precision_15: 0.1321 - val_recall_15: 0.9336 - val_f1_score: 0.1880
Epoch 3/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.3191 - categorical_accuracy: 0.6049 - precision_15: 0.1340 - recall_15: 0.9572 - f1_score: 0.2008 - val_loss: 1.5953 - val_categorical_accuracy: 0.5349 - val_precision_15: 0.1293 - val_recall_15: 0.9399 - val_f1_score: 0.1904
Epoch 4/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.3012 - categorical_accuracy: 0.6148 - precision_15: 0.1332 - recall_15: 0.9573 - f1_score: 0.2066 - val_loss: 1.5855 - val_categorical_accuracy: 0.5036 - val_precision_15: 0.1321 - val_recall_15: 0.9577 - val_f1_score: 0.1989
Epoch 5/50
2584/2584 [==============================] - 40s 16ms/step - loss: 1.2977 - categorical_accuracy: 0.6136 - precision_15: 0.1342 - recall_15: 0.9624 - f1_score: 0.2146 - val_loss: 1.6147 - val_categorical_accuracy: 0.5392 - val_precision_15: 0.1281 - val_recall_15: 0.9359 - val_f1_score: 0.1916
Epoch 6/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2669 - categorical_accuracy: 0.6172 - precision_15: 0.1351 - recall_15: 0.9598 - f1_score: 0.2190 - val_loss: 1.5585 - val_categorical_accuracy: 0.5612 - val_precision_15: 0.1325 - val_recall_15: 0.9426 - val_f1_score: 0.2068
Epoch 7/50
2584/2584 [==============================] - 40s 16ms/step - loss: 1.2566 - categorical_accuracy: 0.6227 - precision_15: 0.1358 - recall_15: 0.9587 - f1_score: 0.2204 - val_loss: 1.5598 - val_categorical_accuracy: 0.5693 - val_precision_15: 0.1284 - val_recall_15: 0.9423 - val_f1_score: 0.2004
Epoch 8/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2395 - categorical_accuracy: 0.6335 - precision_15: 0.1365 - recall_15: 0.9591 - f1_score: 0.2235 - val_loss: 1.6191 - val_categorical_accuracy: 0.5542 - val_precision_15: 0.1312 - val_recall_15: 0.9438 - val_f1_score: 0.2070
Epoch 9/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2426 - categorical_accuracy: 0.6278 - precision_15: 0.1370 - recall_15: 0.9567 - f1_score: 0.2283 - val_loss: 1.5468 - val_categorical_accuracy: 0.5630 - val_precision_15: 0.1290 - val_recall_15: 0.9468 - val_f1_score: 0.1986
Epoch 10/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2338 - categorical_accuracy: 0.6284 - precision_15: 0.1387 - recall_15: 0.9576 - f1_score: 0.2309 - val_loss: 1.5787 - val_categorical_accuracy: 0.5531 - val_precision_15: 0.1337 - val_recall_15: 0.9393 - val_f1_score: 0.2103
Epoch 11/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2236 - categorical_accuracy: 0.6353 - precision_15: 0.1400 - recall_15: 0.9564 - f1_score: 0.2357 - val_loss: 1.5304 - val_categorical_accuracy: 0.5546 - val_precision_15: 0.1327 - val_recall_15: 0.9371 - val_f1_score: 0.2056
Epoch 12/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2183 - categorical_accuracy: 0.6359 - precision_15: 0.1394 - recall_15: 0.9522 - f1_score: 0.2353 - val_loss: 1.5987 - val_categorical_accuracy: 0.5835 - val_precision_15: 0.1344 - val_recall_15: 0.9279 - val_f1_score: 0.2111
Epoch 13/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.2113 - categorical_accuracy: 0.6359 - precision_15: 0.1403 - recall_15: 0.9533 - f1_score: 0.2399 - val_loss: 1.5350 - val_categorical_accuracy: 0.5702 - val_precision_15: 0.1359 - val_recall_15: 0.9258 - val_f1_score: 0.2121
Epoch 14/50
2584/2584 [==============================] - 40s 16ms/step - loss: 1.1995 - categorical_accuracy: 0.6491 - precision_15: 0.1431 - recall_15: 0.9493 - f1_score: 0.2422 - val_loss: 1.5781 - val_categorical_accuracy: 0.5537 - val_precision_15: 0.1358 - val_recall_15: 0.9353 - val_f1_score: 0.2119
Epoch 15/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.1991 - categorical_accuracy: 0.6415 - precision_15: 0.1450 - recall_15: 0.9532 - f1_score: 0.2479 - val_loss: 1.5787 - val_categorical_accuracy: 0.5260 - val_precision_15: 0.1382 - val_recall_15: 0.9332 - val_f1_score: 0.2223
Epoch 16/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.1988 - categorical_accuracy: 0.6390 - precision_15: 0.1431 - recall_15: 0.9460 - f1_score: 0.2488 - val_loss: 1.5611 - val_categorical_accuracy: 0.5885 - val_precision_15: 0.1444 - val_recall_15: 0.9237 - val_f1_score: 0.2219
Epoch 17/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.1979 - categorical_accuracy: 0.6455 - precision_15: 0.1461 - recall_15: 0.9461 - f1_score: 0.2505 - val_loss: 1.5341 - val_categorical_accuracy: 0.5576 - val_precision_15: 0.1422 - val_recall_15: 0.9234 - val_f1_score: 0.2159
Epoch 18/50
2584/2584 [==============================] - 41s 16ms/step - loss: 1.1976 - categorical_accuracy: 0.6458 - precision_15: 0.1463 - recall_15: 0.9480 - f1_score: 0.2511 - val_loss: 1.5409 - val_categorical_accuracy: 0.5722 - val_precision_15: 0.1391 - val_recall_15: 0.9310 - val_f1_score: 0.2205

 1 32 30 0.001 189 48 

Epoch 1/50
2595/2595 [==============================] - 42s 16ms/step - loss: 1.6351 - categorical_accuracy: 0.5271 - precision_16: 0.1210 - recall_16: 0.9008 - f1_score: 0.1575 - val_loss: 1.4816 - val_categorical_accuracy: 0.5737 - val_precision_16: 0.1316 - val_recall_16: 0.9472 - val_f1_score: 0.1846
Epoch 2/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.4010 - categorical_accuracy: 0.5789 - precision_16: 0.1276 - recall_16: 0.9448 - f1_score: 0.1830 - val_loss: 1.4672 - val_categorical_accuracy: 0.5934 - val_precision_16: 0.1252 - val_recall_16: 0.9397 - val_f1_score: 0.1815
Epoch 3/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3500 - categorical_accuracy: 0.5874 - precision_16: 0.1269 - recall_16: 0.9528 - f1_score: 0.1923 - val_loss: 1.3682 - val_categorical_accuracy: 0.6299 - val_precision_16: 0.1284 - val_recall_16: 0.9478 - val_f1_score: 0.1993
Epoch 4/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3316 - categorical_accuracy: 0.5998 - precision_16: 0.1285 - recall_16: 0.9542 - f1_score: 0.2019 - val_loss: 1.3497 - val_categorical_accuracy: 0.5845 - val_precision_16: 0.1258 - val_recall_16: 0.9536 - val_f1_score: 0.1904
Epoch 5/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3054 - categorical_accuracy: 0.6099 - precision_16: 0.1285 - recall_16: 0.9534 - f1_score: 0.2137 - val_loss: 1.3163 - val_categorical_accuracy: 0.6035 - val_precision_16: 0.1308 - val_recall_16: 0.9574 - val_f1_score: 0.1981
Epoch 6/50
2595/2595 [==============================] - 41s 16ms/step - loss: 1.3063 - categorical_accuracy: 0.6032 - precision_16: 0.1295 - recall_16: 0.9528 - f1_score: 0.2129 - val_loss: 1.3809 - val_categorical_accuracy: 0.5670 - val_precision_16: 0.1296 - val_recall_16: 0.9532 - val_f1_score: 0.1989

 2 32 30 0.001 190 47 

Epoch 1/50
2461/2461 [==============================] - 43s 17ms/step - loss: 1.5794 - categorical_accuracy: 0.5463 - precision_17: 0.1325 - recall_17: 0.8958 - f1_score: 0.1613 - val_loss: 2.1933 - val_categorical_accuracy: 0.4792 - val_precision_17: 0.1237 - val_recall_17: 0.8634 - val_f1_score: 0.1544
Epoch 2/50
2461/2461 [==============================] - 42s 17ms/step - loss: 1.3716 - categorical_accuracy: 0.5802 - precision_17: 0.1366 - recall_17: 0.9473 - f1_score: 0.1913 - val_loss: 2.1729 - val_categorical_accuracy: 0.4144 - val_precision_17: 0.1228 - val_recall_17: 0.8954 - val_f1_score: 0.1630
Epoch 3/50
2461/2461 [==============================] - 42s 17ms/step - loss: 1.3306 - categorical_accuracy: 0.5926 - precision_17: 0.1343 - recall_17: 0.9545 - f1_score: 0.2011 - val_loss: 2.2222 - val_categorical_accuracy: 0.4614 - val_precision_17: 0.1213 - val_recall_17: 0.8834 - val_f1_score: 0.1606
Epoch 4/50
2461/2461 [==============================] - 42s 17ms/step - loss: 1.2917 - categorical_accuracy: 0.6034 - precision_17: 0.1344 - recall_17: 0.9544 - f1_score: 0.2001 - val_loss: 2.2971 - val_categorical_accuracy: 0.4337 - val_precision_17: 0.1209 - val_recall_17: 0.8583 - val_f1_score: 0.1629
Epoch 5/50
2461/2461 [==============================] - 42s 17ms/step - loss: 1.2801 - categorical_accuracy: 0.6029 - precision_17: 0.1347 - recall_17: 0.9530 - f1_score: 0.2079 - val_loss: 2.2046 - val_categorical_accuracy: 0.4346 - val_precision_17: 0.1217 - val_recall_17: 0.8913 - val_f1_score: 0.1608

 3 32 30 0.001 190 47 

Epoch 1/50
2627/2627 [==============================] - 38s 14ms/step - loss: 1.6274 - categorical_accuracy: 0.5281 - precision_18: 0.1296 - recall_18: 0.9108 - f1_score: 0.1693 - val_loss: 1.4805 - val_categorical_accuracy: 0.5621 - val_precision_18: 0.1321 - val_recall_18: 0.9371 - val_f1_score: 0.1800
Epoch 2/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.4047 - categorical_accuracy: 0.5666 - precision_18: 0.1321 - recall_18: 0.9595 - f1_score: 0.1974 - val_loss: 1.4197 - val_categorical_accuracy: 0.5590 - val_precision_18: 0.1291 - val_recall_18: 0.9588 - val_f1_score: 0.1855
Epoch 3/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.3468 - categorical_accuracy: 0.5838 - precision_18: 0.1313 - recall_18: 0.9640 - f1_score: 0.2061 - val_loss: 1.4461 - val_categorical_accuracy: 0.5729 - val_precision_18: 0.1362 - val_recall_18: 0.9557 - val_f1_score: 0.1912
Epoch 4/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.3353 - categorical_accuracy: 0.5945 - precision_18: 0.1316 - recall_18: 0.9677 - f1_score: 0.2175 - val_loss: 1.4165 - val_categorical_accuracy: 0.5526 - val_precision_18: 0.1307 - val_recall_18: 0.9478 - val_f1_score: 0.1961
Epoch 5/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.3007 - categorical_accuracy: 0.5987 - precision_18: 0.1329 - recall_18: 0.9668 - f1_score: 0.2182 - val_loss: 1.3040 - val_categorical_accuracy: 0.6136 - val_precision_18: 0.1311 - val_recall_18: 0.9472 - val_f1_score: 0.1923
Epoch 6/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.2928 - categorical_accuracy: 0.6014 - precision_18: 0.1331 - recall_18: 0.9669 - f1_score: 0.2260 - val_loss: 1.3934 - val_categorical_accuracy: 0.5738 - val_precision_18: 0.1363 - val_recall_18: 0.9491 - val_f1_score: 0.2065
Epoch 7/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2819 - categorical_accuracy: 0.6038 - precision_18: 0.1342 - recall_18: 0.9647 - f1_score: 0.2314 - val_loss: 1.3604 - val_categorical_accuracy: 0.5892 - val_precision_18: 0.1353 - val_recall_18: 0.9542 - val_f1_score: 0.2094
Epoch 8/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2631 - categorical_accuracy: 0.6077 - precision_18: 0.1354 - recall_18: 0.9650 - f1_score: 0.2292 - val_loss: 1.3642 - val_categorical_accuracy: 0.5873 - val_precision_18: 0.1355 - val_recall_18: 0.9401 - val_f1_score: 0.2095
Epoch 9/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.2567 - categorical_accuracy: 0.6159 - precision_18: 0.1377 - recall_18: 0.9649 - f1_score: 0.2442 - val_loss: 1.2565 - val_categorical_accuracy: 0.6433 - val_precision_18: 0.1351 - val_recall_18: 0.9396 - val_f1_score: 0.2004
Epoch 10/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2505 - categorical_accuracy: 0.6200 - precision_18: 0.1388 - recall_18: 0.9613 - f1_score: 0.2507 - val_loss: 1.3038 - val_categorical_accuracy: 0.6205 - val_precision_18: 0.1390 - val_recall_18: 0.9478 - val_f1_score: 0.2178
Epoch 11/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.2401 - categorical_accuracy: 0.6214 - precision_18: 0.1409 - recall_18: 0.9629 - f1_score: 0.2576 - val_loss: 1.2882 - val_categorical_accuracy: 0.6416 - val_precision_18: 0.1372 - val_recall_18: 0.9422 - val_f1_score: 0.2150
Epoch 12/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2365 - categorical_accuracy: 0.6223 - precision_18: 0.1412 - recall_18: 0.9613 - f1_score: 0.2533 - val_loss: 1.3872 - val_categorical_accuracy: 0.5587 - val_precision_18: 0.1415 - val_recall_18: 0.9449 - val_f1_score: 0.2233
Epoch 13/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2347 - categorical_accuracy: 0.6236 - precision_18: 0.1416 - recall_18: 0.9594 - f1_score: 0.2712 - val_loss: 1.2800 - val_categorical_accuracy: 0.6224 - val_precision_18: 0.1427 - val_recall_18: 0.9446 - val_f1_score: 0.2213
Epoch 14/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.2331 - categorical_accuracy: 0.6216 - precision_18: 0.1432 - recall_18: 0.9604 - f1_score: 0.2832 - val_loss: 1.3322 - val_categorical_accuracy: 0.5996 - val_precision_18: 0.1431 - val_recall_18: 0.9405 - val_f1_score: 0.2292
Epoch 15/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2302 - categorical_accuracy: 0.6213 - precision_18: 0.1440 - recall_18: 0.9572 - f1_score: 0.2818 - val_loss: 1.2948 - val_categorical_accuracy: 0.6487 - val_precision_18: 0.1477 - val_recall_18: 0.9372 - val_f1_score: 0.2353
Epoch 16/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.2228 - categorical_accuracy: 0.6290 - precision_18: 0.1456 - recall_18: 0.9544 - f1_score: 0.2854 - val_loss: 1.2860 - val_categorical_accuracy: 0.6305 - val_precision_18: 0.1469 - val_recall_18: 0.9349 - val_f1_score: 0.2230
Epoch 17/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2221 - categorical_accuracy: 0.6295 - precision_18: 0.1471 - recall_18: 0.9534 - f1_score: 0.2852 - val_loss: 1.3019 - val_categorical_accuracy: 0.6162 - val_precision_18: 0.1490 - val_recall_18: 0.9297 - val_f1_score: 0.2367
Epoch 18/50
2627/2627 [==============================] - 44s 17ms/step - loss: 1.2179 - categorical_accuracy: 0.6291 - precision_18: 0.1471 - recall_18: 0.9521 - f1_score: 0.2884 - val_loss: 1.3036 - val_categorical_accuracy: 0.6279 - val_precision_18: 0.1452 - val_recall_18: 0.9327 - val_f1_score: 0.2277
Epoch 19/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2185 - categorical_accuracy: 0.6250 - precision_18: 0.1493 - recall_18: 0.9525 - f1_score: 0.2914 - val_loss: 1.3401 - val_categorical_accuracy: 0.5849 - val_precision_18: 0.1456 - val_recall_18: 0.9292 - val_f1_score: 0.2383
Epoch 20/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2179 - categorical_accuracy: 0.6355 - precision_18: 0.1492 - recall_18: 0.9494 - f1_score: 0.2920 - val_loss: 1.2992 - val_categorical_accuracy: 0.5986 - val_precision_18: 0.1476 - val_recall_18: 0.9308 - val_f1_score: 0.2273
Epoch 21/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2101 - categorical_accuracy: 0.6335 - precision_18: 0.1498 - recall_18: 0.9509 - f1_score: 0.2979 - val_loss: 1.2749 - val_categorical_accuracy: 0.6221 - val_precision_18: 0.1501 - val_recall_18: 0.9290 - val_f1_score: 0.2334
Epoch 22/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2166 - categorical_accuracy: 0.6374 - precision_18: 0.1505 - recall_18: 0.9455 - f1_score: 0.2968 - val_loss: 1.2935 - val_categorical_accuracy: 0.6219 - val_precision_18: 0.1502 - val_recall_18: 0.9282 - val_f1_score: 0.2415
Epoch 23/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2167 - categorical_accuracy: 0.6282 - precision_18: 0.1519 - recall_18: 0.9497 - f1_score: 0.3042 - val_loss: 1.2691 - val_categorical_accuracy: 0.6173 - val_precision_18: 0.1478 - val_recall_18: 0.9286 - val_f1_score: 0.2315
Epoch 24/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2109 - categorical_accuracy: 0.6325 - precision_18: 0.1527 - recall_18: 0.9474 - f1_score: 0.3028 - val_loss: 1.2468 - val_categorical_accuracy: 0.6680 - val_precision_18: 0.1524 - val_recall_18: 0.9237 - val_f1_score: 0.2448
Epoch 25/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2109 - categorical_accuracy: 0.6355 - precision_18: 0.1541 - recall_18: 0.9462 - f1_score: 0.3040 - val_loss: 1.2852 - val_categorical_accuracy: 0.6054 - val_precision_18: 0.1510 - val_recall_18: 0.9265 - val_f1_score: 0.2420
Epoch 26/50
2627/2627 [==============================] - 47s 18ms/step - loss: 1.2068 - categorical_accuracy: 0.6319 - precision_18: 0.1546 - recall_18: 0.9466 - f1_score: 0.3088 - val_loss: 1.3771 - val_categorical_accuracy: 0.5661 - val_precision_18: 0.1536 - val_recall_18: 0.9296 - val_f1_score: 0.2535
Epoch 27/50
2627/2627 [==============================] - 46s 18ms/step - loss: 1.2074 - categorical_accuracy: 0.6348 - precision_18: 0.1558 - recall_18: 0.9411 - f1_score: 0.3077 - val_loss: 1.2724 - val_categorical_accuracy: 0.6410 - val_precision_18: 0.1489 - val_recall_18: 0.9140 - val_f1_score: 0.2351
Epoch 28/50
2627/2627 [==============================] - 45s 17ms/step - loss: 1.2041 - categorical_accuracy: 0.6333 - precision_18: 0.1549 - recall_18: 0.9438 - f1_score: 0.3100 - val_loss: 1.2908 - val_categorical_accuracy: 0.6149 - val_precision_18: 0.1533 - val_recall_18: 0.9117 - val_f1_score: 0.2297
Epoch 29/50
2627/2627 [==============================] - 38s 14ms/step - loss: 1.2078 - categorical_accuracy: 0.6387 - precision_18: 0.1572 - recall_18: 0.9392 - f1_score: 0.3087 - val_loss: 1.3062 - val_categorical_accuracy: 0.6126 - val_precision_18: 0.1587 - val_recall_18: 0.9262 - val_f1_score: 0.2533

 4 32 30 0.001 190 47 

Epoch 1/50
2625/2625 [==============================] - 47s 18ms/step - loss: 1.6425 - categorical_accuracy: 0.5241 - precision_19: 0.1205 - recall_19: 0.9014 - f1_score: 0.1576 - val_loss: 1.3396 - val_categorical_accuracy: 0.5975 - val_precision_19: 0.1318 - val_recall_19: 0.9421 - val_f1_score: 0.1671
Epoch 2/50
2625/2625 [==============================] - 46s 17ms/step - loss: 1.4083 - categorical_accuracy: 0.5709 - precision_19: 0.1255 - recall_19: 0.9493 - f1_score: 0.1824 - val_loss: 1.3520 - val_categorical_accuracy: 0.6126 - val_precision_19: 0.1279 - val_recall_19: 0.9548 - val_f1_score: 0.1743
Epoch 3/50
2625/2625 [==============================] - 47s 18ms/step - loss: 1.3606 - categorical_accuracy: 0.5813 - precision_19: 0.1269 - recall_19: 0.9560 - f1_score: 0.1940 - val_loss: 1.2969 - val_categorical_accuracy: 0.6207 - val_precision_19: 0.1299 - val_recall_19: 0.9483 - val_f1_score: 0.1823
Epoch 4/50
2625/2625 [==============================] - 46s 18ms/step - loss: 1.3428 - categorical_accuracy: 0.5972 - precision_19: 0.1278 - recall_19: 0.9561 - f1_score: 0.2066 - val_loss: 1.2998 - val_categorical_accuracy: 0.6274 - val_precision_19: 0.1306 - val_recall_19: 0.9674 - val_f1_score: 0.1822
Epoch 5/50
2625/2625 [==============================] - 46s 17ms/step - loss: 1.3272 - categorical_accuracy: 0.5935 - precision_19: 0.1284 - recall_19: 0.9560 - f1_score: 0.2023 - val_loss: 1.3752 - val_categorical_accuracy: 0.6191 - val_precision_19: 0.1242 - val_recall_19: 0.9552 - val_f1_score: 0.1747
Epoch 6/50
2625/2625 [==============================] - 46s 18ms/step - loss: 1.3058 - categorical_accuracy: 0.6013 - precision_19: 0.1288 - recall_19: 0.9567 - f1_score: 0.2094 - val_loss: 1.2324 - val_categorical_accuracy: 0.6409 - val_precision_19: 0.1298 - val_recall_19: 0.9687 - val_f1_score: 0.1796

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

 0 64 15 0.0001 189 48 

Epoch 1/50
1367/1367 [==============================] - 46s 33ms/step - loss: 2.5924 - categorical_accuracy: 0.2929 - precision_20: 0.0877 - recall_20: 0.6630 - f1_score: 0.1285 - val_loss: 2.3139 - val_categorical_accuracy: 0.3560 - val_precision_20: 0.0988 - val_recall_20: 0.7370 - val_f1_score: 0.1348
Epoch 2/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.9849 - categorical_accuracy: 0.4160 - precision_20: 0.1113 - recall_20: 0.8079 - f1_score: 0.1506 - val_loss: 2.0233 - val_categorical_accuracy: 0.4336 - val_precision_20: 0.1110 - val_recall_20: 0.8030 - val_f1_score: 0.1411
Epoch 3/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.7452 - categorical_accuracy: 0.4860 - precision_20: 0.1226 - recall_20: 0.8550 - f1_score: 0.1580 - val_loss: 1.8553 - val_categorical_accuracy: 0.4747 - val_precision_20: 0.1212 - val_recall_20: 0.8446 - val_f1_score: 0.1474
Epoch 4/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.6037 - categorical_accuracy: 0.5282 - precision_20: 0.1312 - recall_20: 0.8829 - f1_score: 0.1639 - val_loss: 1.7548 - val_categorical_accuracy: 0.4993 - val_precision_20: 0.1265 - val_recall_20: 0.8590 - val_f1_score: 0.1489
Epoch 5/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.5177 - categorical_accuracy: 0.5495 - precision_20: 0.1357 - recall_20: 0.8960 - f1_score: 0.1667 - val_loss: 1.6997 - val_categorical_accuracy: 0.5216 - val_precision_20: 0.1302 - val_recall_20: 0.8676 - val_f1_score: 0.1513
Epoch 6/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.4632 - categorical_accuracy: 0.5716 - precision_20: 0.1375 - recall_20: 0.9067 - f1_score: 0.1698 - val_loss: 1.6620 - val_categorical_accuracy: 0.5260 - val_precision_20: 0.1305 - val_recall_20: 0.8742 - val_f1_score: 0.1538
Epoch 7/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.4256 - categorical_accuracy: 0.5758 - precision_20: 0.1374 - recall_20: 0.9140 - f1_score: 0.1726 - val_loss: 1.6396 - val_categorical_accuracy: 0.5355 - val_precision_20: 0.1306 - val_recall_20: 0.8781 - val_f1_score: 0.1562
Epoch 8/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3991 - categorical_accuracy: 0.5874 - precision_20: 0.1373 - recall_20: 0.9207 - f1_score: 0.1751 - val_loss: 1.6238 - val_categorical_accuracy: 0.5388 - val_precision_20: 0.1304 - val_recall_20: 0.8825 - val_f1_score: 0.1585
Epoch 9/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3774 - categorical_accuracy: 0.5930 - precision_20: 0.1372 - recall_20: 0.9256 - f1_score: 0.1778 - val_loss: 1.6143 - val_categorical_accuracy: 0.5416 - val_precision_20: 0.1312 - val_recall_20: 0.8883 - val_f1_score: 0.1600
Epoch 10/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3618 - categorical_accuracy: 0.5920 - precision_20: 0.1380 - recall_20: 0.9306 - f1_score: 0.1805 - val_loss: 1.6081 - val_categorical_accuracy: 0.5520 - val_precision_20: 0.1311 - val_recall_20: 0.8908 - val_f1_score: 0.1623
Epoch 11/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3479 - categorical_accuracy: 0.6008 - precision_20: 0.1379 - recall_20: 0.9332 - f1_score: 0.1834 - val_loss: 1.5988 - val_categorical_accuracy: 0.5544 - val_precision_20: 0.1314 - val_recall_20: 0.8950 - val_f1_score: 0.1637
Epoch 12/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3354 - categorical_accuracy: 0.6042 - precision_20: 0.1386 - recall_20: 0.9369 - f1_score: 0.1868 - val_loss: 1.6010 - val_categorical_accuracy: 0.5499 - val_precision_20: 0.1312 - val_recall_20: 0.8982 - val_f1_score: 0.1652
Epoch 13/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3268 - categorical_accuracy: 0.6046 - precision_20: 0.1378 - recall_20: 0.9384 - f1_score: 0.1872 - val_loss: 1.6023 - val_categorical_accuracy: 0.5502 - val_precision_20: 0.1321 - val_recall_20: 0.9020 - val_f1_score: 0.1681
Epoch 14/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3173 - categorical_accuracy: 0.6065 - precision_20: 0.1382 - recall_20: 0.9406 - f1_score: 0.1898 - val_loss: 1.5910 - val_categorical_accuracy: 0.5587 - val_precision_20: 0.1318 - val_recall_20: 0.9036 - val_f1_score: 0.1676
Epoch 15/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3098 - categorical_accuracy: 0.6071 - precision_20: 0.1384 - recall_20: 0.9427 - f1_score: 0.1918 - val_loss: 1.6030 - val_categorical_accuracy: 0.5512 - val_precision_20: 0.1316 - val_recall_20: 0.9053 - val_f1_score: 0.1694
Epoch 16/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3056 - categorical_accuracy: 0.6070 - precision_20: 0.1380 - recall_20: 0.9436 - f1_score: 0.1930 - val_loss: 1.5983 - val_categorical_accuracy: 0.5620 - val_precision_20: 0.1310 - val_recall_20: 0.9056 - val_f1_score: 0.1700
Epoch 17/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2992 - categorical_accuracy: 0.6147 - precision_20: 0.1384 - recall_20: 0.9450 - f1_score: 0.1934 - val_loss: 1.5963 - val_categorical_accuracy: 0.5647 - val_precision_20: 0.1314 - val_recall_20: 0.9077 - val_f1_score: 0.1714
Epoch 18/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2947 - categorical_accuracy: 0.6140 - precision_20: 0.1382 - recall_20: 0.9448 - f1_score: 0.1944 - val_loss: 1.5936 - val_categorical_accuracy: 0.5641 - val_precision_20: 0.1313 - val_recall_20: 0.9090 - val_f1_score: 0.1726
Epoch 19/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2885 - categorical_accuracy: 0.6174 - precision_20: 0.1380 - recall_20: 0.9465 - f1_score: 0.1950 - val_loss: 1.6004 - val_categorical_accuracy: 0.5658 - val_precision_20: 0.1314 - val_recall_20: 0.9100 - val_f1_score: 0.1729
Epoch 20/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2858 - categorical_accuracy: 0.6163 - precision_20: 0.1383 - recall_20: 0.9463 - f1_score: 0.1964 - val_loss: 1.5978 - val_categorical_accuracy: 0.5662 - val_precision_20: 0.1316 - val_recall_20: 0.9123 - val_f1_score: 0.1740
Epoch 21/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2836 - categorical_accuracy: 0.6179 - precision_20: 0.1382 - recall_20: 0.9473 - f1_score: 0.1960 - val_loss: 1.5993 - val_categorical_accuracy: 0.5715 - val_precision_20: 0.1318 - val_recall_20: 0.9131 - val_f1_score: 0.1751
Epoch 22/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2792 - categorical_accuracy: 0.6195 - precision_20: 0.1386 - recall_20: 0.9493 - f1_score: 0.1982 - val_loss: 1.5980 - val_categorical_accuracy: 0.5735 - val_precision_20: 0.1315 - val_recall_20: 0.9159 - val_f1_score: 0.1752
Epoch 23/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2756 - categorical_accuracy: 0.6234 - precision_20: 0.1383 - recall_20: 0.9500 - f1_score: 0.1977 - val_loss: 1.6032 - val_categorical_accuracy: 0.5664 - val_precision_20: 0.1316 - val_recall_20: 0.9176 - val_f1_score: 0.1770
Epoch 24/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2743 - categorical_accuracy: 0.6209 - precision_20: 0.1379 - recall_20: 0.9495 - f1_score: 0.1990 - val_loss: 1.6018 - val_categorical_accuracy: 0.5725 - val_precision_20: 0.1315 - val_recall_20: 0.9196 - val_f1_score: 0.1760
Epoch 25/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2705 - categorical_accuracy: 0.6215 - precision_20: 0.1381 - recall_20: 0.9514 - f1_score: 0.1990 - val_loss: 1.6064 - val_categorical_accuracy: 0.5759 - val_precision_20: 0.1313 - val_recall_20: 0.9194 - val_f1_score: 0.1762
Epoch 26/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2667 - categorical_accuracy: 0.6235 - precision_20: 0.1382 - recall_20: 0.9515 - f1_score: 0.1996 - val_loss: 1.6037 - val_categorical_accuracy: 0.5767 - val_precision_20: 0.1313 - val_recall_20: 0.9212 - val_f1_score: 0.1765

 1 64 15 0.0001 189 48 

Epoch 1/50
1371/1371 [==============================] - 19s 13ms/step - loss: 2.4716 - categorical_accuracy: 0.3253 - precision_21: 0.0916 - recall_21: 0.7281 - f1_score: 0.1344 - val_loss: 2.1138 - val_categorical_accuracy: 0.4188 - val_precision_21: 0.0964 - val_recall_21: 0.7630 - val_f1_score: 0.1371
Epoch 2/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.9719 - categorical_accuracy: 0.4451 - precision_21: 0.1075 - recall_21: 0.8107 - f1_score: 0.1473 - val_loss: 1.8813 - val_categorical_accuracy: 0.4458 - val_precision_21: 0.1068 - val_recall_21: 0.8036 - val_f1_score: 0.1422
Epoch 3/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.7681 - categorical_accuracy: 0.4830 - precision_21: 0.1189 - recall_21: 0.8565 - f1_score: 0.1539 - val_loss: 1.7289 - val_categorical_accuracy: 0.4710 - val_precision_21: 0.1171 - val_recall_21: 0.8545 - val_f1_score: 0.1491
Epoch 4/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.6340 - categorical_accuracy: 0.5196 - precision_21: 0.1256 - recall_21: 0.8812 - f1_score: 0.1578 - val_loss: 1.6272 - val_categorical_accuracy: 0.5018 - val_precision_21: 0.1227 - val_recall_21: 0.8713 - val_f1_score: 0.1554
Epoch 5/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.5488 - categorical_accuracy: 0.5456 - precision_21: 0.1275 - recall_21: 0.8934 - f1_score: 0.1620 - val_loss: 1.5659 - val_categorical_accuracy: 0.5214 - val_precision_21: 0.1244 - val_recall_21: 0.8861 - val_f1_score: 0.1605
Epoch 6/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.4928 - categorical_accuracy: 0.5609 - precision_21: 0.1297 - recall_21: 0.9049 - f1_score: 0.1678 - val_loss: 1.5305 - val_categorical_accuracy: 0.5374 - val_precision_21: 0.1257 - val_recall_21: 0.8951 - val_f1_score: 0.1647
Epoch 7/50
1371/1371 [==============================] - 16s 12ms/step - loss: 1.4576 - categorical_accuracy: 0.5730 - precision_21: 0.1315 - recall_21: 0.9127 - f1_score: 0.1724 - val_loss: 1.4958 - val_categorical_accuracy: 0.5491 - val_precision_21: 0.1271 - val_recall_21: 0.9031 - val_f1_score: 0.1698
Epoch 8/50
1371/1371 [==============================] - 15s 11ms/step - loss: 1.4271 - categorical_accuracy: 0.5837 - precision_21: 0.1318 - recall_21: 0.9176 - f1_score: 0.1752 - val_loss: 1.4727 - val_categorical_accuracy: 0.5582 - val_precision_21: 0.1282 - val_recall_21: 0.9106 - val_f1_score: 0.1746
Epoch 9/50
1371/1371 [==============================] - 17s 12ms/step - loss: 1.4069 - categorical_accuracy: 0.5856 - precision_21: 0.1324 - recall_21: 0.9223 - f1_score: 0.1785 - val_loss: 1.4498 - val_categorical_accuracy: 0.5600 - val_precision_21: 0.1291 - val_recall_21: 0.9144 - val_f1_score: 0.1768
Epoch 10/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3896 - categorical_accuracy: 0.5891 - precision_21: 0.1335 - recall_21: 0.9272 - f1_score: 0.1826 - val_loss: 1.4390 - val_categorical_accuracy: 0.5665 - val_precision_21: 0.1298 - val_recall_21: 0.9183 - val_f1_score: 0.1794
Epoch 11/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3752 - categorical_accuracy: 0.5970 - precision_21: 0.1333 - recall_21: 0.9299 - f1_score: 0.1822 - val_loss: 1.4251 - val_categorical_accuracy: 0.5607 - val_precision_21: 0.1320 - val_recall_21: 0.9250 - val_f1_score: 0.1847
Epoch 12/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3656 - categorical_accuracy: 0.5927 - precision_21: 0.1340 - recall_21: 0.9341 - f1_score: 0.1888 - val_loss: 1.4076 - val_categorical_accuracy: 0.5736 - val_precision_21: 0.1325 - val_recall_21: 0.9270 - val_f1_score: 0.1878
Epoch 13/50
1371/1371 [==============================] - 17s 13ms/step - loss: 1.3555 - categorical_accuracy: 0.5980 - precision_21: 0.1344 - recall_21: 0.9353 - f1_score: 0.1924 - val_loss: 1.4001 - val_categorical_accuracy: 0.5792 - val_precision_21: 0.1318 - val_recall_21: 0.9305 - val_f1_score: 0.1860
Epoch 14/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3455 - categorical_accuracy: 0.5998 - precision_21: 0.1344 - recall_21: 0.9377 - f1_score: 0.1929 - val_loss: 1.3941 - val_categorical_accuracy: 0.5777 - val_precision_21: 0.1327 - val_recall_21: 0.9321 - val_f1_score: 0.1883
Epoch 15/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3386 - categorical_accuracy: 0.6035 - precision_21: 0.1346 - recall_21: 0.9388 - f1_score: 0.2054 - val_loss: 1.3915 - val_categorical_accuracy: 0.5813 - val_precision_21: 0.1325 - val_recall_21: 0.9348 - val_f1_score: 0.1900
Epoch 16/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3322 - categorical_accuracy: 0.6051 - precision_21: 0.1342 - recall_21: 0.9397 - f1_score: 0.2014 - val_loss: 1.3810 - val_categorical_accuracy: 0.5858 - val_precision_21: 0.1331 - val_recall_21: 0.9366 - val_f1_score: 0.1921
Epoch 17/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3263 - categorical_accuracy: 0.6085 - precision_21: 0.1349 - recall_21: 0.9420 - f1_score: 0.2052 - val_loss: 1.3743 - val_categorical_accuracy: 0.5908 - val_precision_21: 0.1319 - val_recall_21: 0.9366 - val_f1_score: 0.1917
Epoch 18/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3210 - categorical_accuracy: 0.6094 - precision_21: 0.1340 - recall_21: 0.9428 - f1_score: 0.2078 - val_loss: 1.3737 - val_categorical_accuracy: 0.5834 - val_precision_21: 0.1330 - val_recall_21: 0.9379 - val_f1_score: 0.1937
Epoch 19/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3177 - categorical_accuracy: 0.6154 - precision_21: 0.1339 - recall_21: 0.9422 - f1_score: 0.2078 - val_loss: 1.3662 - val_categorical_accuracy: 0.5897 - val_precision_21: 0.1332 - val_recall_21: 0.9413 - val_f1_score: 0.1955
Epoch 20/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3118 - categorical_accuracy: 0.6166 - precision_21: 0.1345 - recall_21: 0.9440 - f1_score: 0.2103 - val_loss: 1.3623 - val_categorical_accuracy: 0.5874 - val_precision_21: 0.1324 - val_recall_21: 0.9407 - val_f1_score: 0.1945
Epoch 21/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3101 - categorical_accuracy: 0.6133 - precision_21: 0.1333 - recall_21: 0.9441 - f1_score: 0.2119 - val_loss: 1.3612 - val_categorical_accuracy: 0.5865 - val_precision_21: 0.1331 - val_recall_21: 0.9418 - val_f1_score: 0.1953
Epoch 22/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3058 - categorical_accuracy: 0.6183 - precision_21: 0.1337 - recall_21: 0.9445 - f1_score: 0.2106 - val_loss: 1.3586 - val_categorical_accuracy: 0.5873 - val_precision_21: 0.1332 - val_recall_21: 0.9438 - val_f1_score: 0.1954

 2 64 15 0.0001 190 47 

Epoch 1/50
1304/1304 [==============================] - 19s 14ms/step - loss: 2.4586 - categorical_accuracy: 0.3377 - precision_22: 0.0979 - recall_22: 0.7183 - f1_score: 0.1366 - val_loss: 2.3886 - val_categorical_accuracy: 0.3301 - val_precision_22: 0.0971 - val_recall_22: 0.7057 - val_f1_score: 0.1317
Epoch 2/50
1304/1304 [==============================] - 18s 14ms/step - loss: 1.9269 - categorical_accuracy: 0.4644 - precision_22: 0.1166 - recall_22: 0.8035 - f1_score: 0.1418 - val_loss: 2.1984 - val_categorical_accuracy: 0.3800 - val_precision_22: 0.1073 - val_recall_22: 0.7458 - val_f1_score: 0.1370
Epoch 3/50
1304/1304 [==============================] - 18s 14ms/step - loss: 1.7148 - categorical_accuracy: 0.5137 - precision_22: 0.1299 - recall_22: 0.8563 - f1_score: 0.1490 - val_loss: 2.1056 - val_categorical_accuracy: 0.4089 - val_precision_22: 0.1164 - val_recall_22: 0.7689 - val_f1_score: 0.1425
Epoch 4/50
1304/1304 [==============================] - 18s 14ms/step - loss: 1.5816 - categorical_accuracy: 0.5439 - precision_22: 0.1395 - recall_22: 0.8882 - f1_score: 0.1541 - val_loss: 2.0680 - val_categorical_accuracy: 0.4260 - val_precision_22: 0.1213 - val_recall_22: 0.7805 - val_f1_score: 0.1443
Epoch 5/50
1304/1304 [==============================] - 15s 11ms/step - loss: 1.4986 - categorical_accuracy: 0.5664 - precision_22: 0.1434 - recall_22: 0.9018 - f1_score: 0.1603 - val_loss: 2.0538 - val_categorical_accuracy: 0.4398 - val_precision_22: 0.1238 - val_recall_22: 0.7994 - val_f1_score: 0.1472
Epoch 6/50
1304/1304 [==============================] - 15s 12ms/step - loss: 1.4469 - categorical_accuracy: 0.5795 - precision_22: 0.1427 - recall_22: 0.9120 - f1_score: 0.1635 - val_loss: 2.0530 - val_categorical_accuracy: 0.4451 - val_precision_22: 0.1265 - val_recall_22: 0.8214 - val_f1_score: 0.1512
Epoch 7/50
1304/1304 [==============================] - 15s 11ms/step - loss: 1.4089 - categorical_accuracy: 0.5848 - precision_22: 0.1438 - recall_22: 0.9179 - f1_score: 0.1684 - val_loss: 2.0430 - val_categorical_accuracy: 0.4561 - val_precision_22: 0.1270 - val_recall_22: 0.8372 - val_f1_score: 0.1536
Epoch 8/50
1304/1304 [==============================] - 15s 11ms/step - loss: 1.3840 - categorical_accuracy: 0.5855 - precision_22: 0.1419 - recall_22: 0.9268 - f1_score: 0.1716 - val_loss: 2.0567 - val_categorical_accuracy: 0.4577 - val_precision_22: 0.1275 - val_recall_22: 0.8484 - val_f1_score: 0.1550
Epoch 9/50
1304/1304 [==============================] - 15s 11ms/step - loss: 1.3640 - categorical_accuracy: 0.5952 - precision_22: 0.1413 - recall_22: 0.9312 - f1_score: 0.1767 - val_loss: 2.0595 - val_categorical_accuracy: 0.4543 - val_precision_22: 0.1278 - val_recall_22: 0.8599 - val_f1_score: 0.1559
Epoch 10/50
1304/1304 [==============================] - 15s 11ms/step - loss: 1.3457 - categorical_accuracy: 0.5935 - precision_22: 0.1402 - recall_22: 0.9368 - f1_score: 0.1787 - val_loss: 2.0619 - val_categorical_accuracy: 0.4632 - val_precision_22: 0.1269 - val_recall_22: 0.8668 - val_f1_score: 0.1559
Epoch 11/50
1304/1304 [==============================] - 16s 12ms/step - loss: 1.3332 - categorical_accuracy: 0.6019 - precision_22: 0.1400 - recall_22: 0.9406 - f1_score: 0.1829 - val_loss: 2.0702 - val_categorical_accuracy: 0.4633 - val_precision_22: 0.1270 - val_recall_22: 0.8718 - val_f1_score: 0.1577
Epoch 12/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.3222 - categorical_accuracy: 0.6036 - precision_22: 0.1384 - recall_22: 0.9436 - f1_score: 0.1859 - val_loss: 2.0896 - val_categorical_accuracy: 0.4641 - val_precision_22: 0.1278 - val_recall_22: 0.8780 - val_f1_score: 0.1568
Epoch 13/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.3136 - categorical_accuracy: 0.6057 - precision_22: 0.1380 - recall_22: 0.9462 - f1_score: 0.1880 - val_loss: 2.0810 - val_categorical_accuracy: 0.4644 - val_precision_22: 0.1274 - val_recall_22: 0.8803 - val_f1_score: 0.1590
Epoch 14/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.3043 - categorical_accuracy: 0.6101 - precision_22: 0.1379 - recall_22: 0.9485 - f1_score: 0.1881 - val_loss: 2.0983 - val_categorical_accuracy: 0.4657 - val_precision_22: 0.1266 - val_recall_22: 0.8804 - val_f1_score: 0.1580
Epoch 15/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.3000 - categorical_accuracy: 0.6105 - precision_22: 0.1380 - recall_22: 0.9498 - f1_score: 0.1911 - val_loss: 2.1009 - val_categorical_accuracy: 0.4749 - val_precision_22: 0.1259 - val_recall_22: 0.8832 - val_f1_score: 0.1582
Epoch 16/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.2928 - categorical_accuracy: 0.6131 - precision_22: 0.1366 - recall_22: 0.9523 - f1_score: 0.1934 - val_loss: 2.1195 - val_categorical_accuracy: 0.4737 - val_precision_22: 0.1249 - val_recall_22: 0.8870 - val_f1_score: 0.1588

 3 64 15 0.0001 190 47 

Epoch 1/50
1387/1387 [==============================] - 19s 13ms/step - loss: 2.6039 - categorical_accuracy: 0.2928 - precision_23: 0.0848 - recall_23: 0.6722 - f1_score: 0.1246 - val_loss: 2.1040 - val_categorical_accuracy: 0.4028 - val_precision_23: 0.1040 - val_recall_23: 0.7876 - val_f1_score: 0.1408
Epoch 2/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.9701 - categorical_accuracy: 0.4323 - precision_23: 0.1068 - recall_23: 0.8150 - f1_score: 0.1457 - val_loss: 1.7780 - val_categorical_accuracy: 0.5170 - val_precision_23: 0.1114 - val_recall_23: 0.8542 - val_f1_score: 0.1469
Epoch 3/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.7457 - categorical_accuracy: 0.4991 - precision_23: 0.1146 - recall_23: 0.8617 - f1_score: 0.1506 - val_loss: 1.6334 - val_categorical_accuracy: 0.5547 - val_precision_23: 0.1181 - val_recall_23: 0.8766 - val_f1_score: 0.1484
Epoch 4/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.6182 - categorical_accuracy: 0.5282 - precision_23: 0.1198 - recall_23: 0.8818 - f1_score: 0.1531 - val_loss: 1.5422 - val_categorical_accuracy: 0.5757 - val_precision_23: 0.1233 - val_recall_23: 0.8914 - val_f1_score: 0.1520
Epoch 5/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.5380 - categorical_accuracy: 0.5495 - precision_23: 0.1241 - recall_23: 0.8979 - f1_score: 0.1575 - val_loss: 1.4905 - val_categorical_accuracy: 0.5826 - val_precision_23: 0.1239 - val_recall_23: 0.9012 - val_f1_score: 0.1540
Epoch 6/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.4840 - categorical_accuracy: 0.5629 - precision_23: 0.1237 - recall_23: 0.9111 - f1_score: 0.1601 - val_loss: 1.4527 - val_categorical_accuracy: 0.5891 - val_precision_23: 0.1246 - val_recall_23: 0.9082 - val_f1_score: 0.1560
Epoch 7/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.4476 - categorical_accuracy: 0.5690 - precision_23: 0.1253 - recall_23: 0.9202 - f1_score: 0.1644 - val_loss: 1.4354 - val_categorical_accuracy: 0.5939 - val_precision_23: 0.1236 - val_recall_23: 0.9124 - val_f1_score: 0.1599
Epoch 8/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.4188 - categorical_accuracy: 0.5768 - precision_23: 0.1249 - recall_23: 0.9306 - f1_score: 0.1665 - val_loss: 1.4261 - val_categorical_accuracy: 0.5924 - val_precision_23: 0.1240 - val_recall_23: 0.9198 - val_f1_score: 0.1639
Epoch 9/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3998 - categorical_accuracy: 0.5787 - precision_23: 0.1254 - recall_23: 0.9351 - f1_score: 0.1715 - val_loss: 1.4137 - val_categorical_accuracy: 0.5929 - val_precision_23: 0.1244 - val_recall_23: 0.9245 - val_f1_score: 0.1673
Epoch 10/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3814 - categorical_accuracy: 0.5812 - precision_23: 0.1258 - recall_23: 0.9401 - f1_score: 0.1716 - val_loss: 1.4023 - val_categorical_accuracy: 0.5981 - val_precision_23: 0.1251 - val_recall_23: 0.9272 - val_f1_score: 0.1718
Epoch 11/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3672 - categorical_accuracy: 0.5933 - precision_23: 0.1259 - recall_23: 0.9449 - f1_score: 0.1793 - val_loss: 1.3959 - val_categorical_accuracy: 0.5969 - val_precision_23: 0.1252 - val_recall_23: 0.9284 - val_f1_score: 0.1741
Epoch 12/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3581 - categorical_accuracy: 0.5903 - precision_23: 0.1271 - recall_23: 0.9481 - f1_score: 0.1806 - val_loss: 1.3860 - val_categorical_accuracy: 0.6019 - val_precision_23: 0.1247 - val_recall_23: 0.9299 - val_f1_score: 0.1756
Epoch 13/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3486 - categorical_accuracy: 0.5952 - precision_23: 0.1272 - recall_23: 0.9501 - f1_score: 0.1854 - val_loss: 1.3823 - val_categorical_accuracy: 0.6080 - val_precision_23: 0.1247 - val_recall_23: 0.9306 - val_f1_score: 0.1763
Epoch 14/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3393 - categorical_accuracy: 0.5965 - precision_23: 0.1274 - recall_23: 0.9535 - f1_score: 0.1871 - val_loss: 1.3774 - val_categorical_accuracy: 0.6112 - val_precision_23: 0.1252 - val_recall_23: 0.9312 - val_f1_score: 0.1776
Epoch 15/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3326 - categorical_accuracy: 0.6001 - precision_23: 0.1275 - recall_23: 0.9539 - f1_score: 0.1890 - val_loss: 1.3691 - val_categorical_accuracy: 0.6143 - val_precision_23: 0.1258 - val_recall_23: 0.9328 - val_f1_score: 0.1799
Epoch 16/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3266 - categorical_accuracy: 0.5972 - precision_23: 0.1279 - recall_23: 0.9567 - f1_score: 0.1900 - val_loss: 1.3711 - val_categorical_accuracy: 0.6080 - val_precision_23: 0.1260 - val_recall_23: 0.9334 - val_f1_score: 0.1810
Epoch 17/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3206 - categorical_accuracy: 0.5987 - precision_23: 0.1283 - recall_23: 0.9571 - f1_score: 0.1909 - val_loss: 1.3655 - val_categorical_accuracy: 0.6157 - val_precision_23: 0.1262 - val_recall_23: 0.9341 - val_f1_score: 0.1824
Epoch 18/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3170 - categorical_accuracy: 0.6048 - precision_23: 0.1282 - recall_23: 0.9592 - f1_score: 0.1923 - val_loss: 1.3635 - val_categorical_accuracy: 0.6174 - val_precision_23: 0.1259 - val_recall_23: 0.9347 - val_f1_score: 0.1835
Epoch 19/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3116 - categorical_accuracy: 0.6039 - precision_23: 0.1284 - recall_23: 0.9596 - f1_score: 0.1926 - val_loss: 1.3698 - val_categorical_accuracy: 0.6111 - val_precision_23: 0.1266 - val_recall_23: 0.9352 - val_f1_score: 0.1852
Epoch 20/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3063 - categorical_accuracy: 0.6074 - precision_23: 0.1288 - recall_23: 0.9624 - f1_score: 0.1954 - val_loss: 1.3714 - val_categorical_accuracy: 0.6071 - val_precision_23: 0.1263 - val_recall_23: 0.9362 - val_f1_score: 0.1853
Epoch 21/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3046 - categorical_accuracy: 0.6083 - precision_23: 0.1285 - recall_23: 0.9604 - f1_score: 0.1955 - val_loss: 1.3642 - val_categorical_accuracy: 0.6172 - val_precision_23: 0.1257 - val_recall_23: 0.9369 - val_f1_score: 0.1840
Epoch 22/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.2983 - categorical_accuracy: 0.6100 - precision_23: 0.1288 - recall_23: 0.9627 - f1_score: 0.1961 - val_loss: 1.3610 - val_categorical_accuracy: 0.6207 - val_precision_23: 0.1260 - val_recall_23: 0.9363 - val_f1_score: 0.1850
Epoch 23/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.2983 - categorical_accuracy: 0.6082 - precision_23: 0.1286 - recall_23: 0.9626 - f1_score: 0.1962 - val_loss: 1.3569 - val_categorical_accuracy: 0.6158 - val_precision_23: 0.1260 - val_recall_23: 0.9363 - val_f1_score: 0.1851

 4 64 15 0.0001 190 47 

Epoch 1/50
1387/1387 [==============================] - 19s 13ms/step - loss: 2.6656 - categorical_accuracy: 0.2583 - precision_24: 0.0840 - recall_24: 0.6897 - f1_score: 0.1268 - val_loss: 2.0798 - val_categorical_accuracy: 0.4219 - val_precision_24: 0.1053 - val_recall_24: 0.8052 - val_f1_score: 0.1410
Epoch 2/50
1387/1387 [==============================] - 17s 12ms/step - loss: 2.0125 - categorical_accuracy: 0.4298 - precision_24: 0.1055 - recall_24: 0.8108 - f1_score: 0.1455 - val_loss: 1.7794 - val_categorical_accuracy: 0.5001 - val_precision_24: 0.1159 - val_recall_24: 0.8431 - val_f1_score: 0.1421
Epoch 3/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.7916 - categorical_accuracy: 0.4882 - precision_24: 0.1173 - recall_24: 0.8445 - f1_score: 0.1486 - val_loss: 1.6377 - val_categorical_accuracy: 0.5345 - val_precision_24: 0.1282 - val_recall_24: 0.8871 - val_f1_score: 0.1496
Epoch 4/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.6602 - categorical_accuracy: 0.5243 - precision_24: 0.1244 - recall_24: 0.8734 - f1_score: 0.1541 - val_loss: 1.5492 - val_categorical_accuracy: 0.5545 - val_precision_24: 0.1296 - val_recall_24: 0.9023 - val_f1_score: 0.1527
Epoch 5/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.5749 - categorical_accuracy: 0.5477 - precision_24: 0.1253 - recall_24: 0.8912 - f1_score: 0.1577 - val_loss: 1.4883 - val_categorical_accuracy: 0.5648 - val_precision_24: 0.1295 - val_recall_24: 0.9134 - val_f1_score: 0.1550
Epoch 6/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.5195 - categorical_accuracy: 0.5598 - precision_24: 0.1263 - recall_24: 0.9065 - f1_score: 0.1616 - val_loss: 1.4487 - val_categorical_accuracy: 0.5735 - val_precision_24: 0.1307 - val_recall_24: 0.9208 - val_f1_score: 0.1587
Epoch 7/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.4800 - categorical_accuracy: 0.5674 - precision_24: 0.1282 - recall_24: 0.9167 - f1_score: 0.1662 - val_loss: 1.4220 - val_categorical_accuracy: 0.5790 - val_precision_24: 0.1300 - val_recall_24: 0.9258 - val_f1_score: 0.1618
Epoch 8/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.4472 - categorical_accuracy: 0.5740 - precision_24: 0.1284 - recall_24: 0.9257 - f1_score: 0.1720 - val_loss: 1.3991 - val_categorical_accuracy: 0.5872 - val_precision_24: 0.1306 - val_recall_24: 0.9292 - val_f1_score: 0.1644
Epoch 9/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.4256 - categorical_accuracy: 0.5789 - precision_24: 0.1287 - recall_24: 0.9306 - f1_score: 0.1744 - val_loss: 1.3859 - val_categorical_accuracy: 0.5877 - val_precision_24: 0.1310 - val_recall_24: 0.9317 - val_f1_score: 0.1649
Epoch 10/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.4054 - categorical_accuracy: 0.5803 - precision_24: 0.1295 - recall_24: 0.9353 - f1_score: 0.1798 - val_loss: 1.3812 - val_categorical_accuracy: 0.5867 - val_precision_24: 0.1312 - val_recall_24: 0.9347 - val_f1_score: 0.1675
Epoch 11/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3923 - categorical_accuracy: 0.5860 - precision_24: 0.1296 - recall_24: 0.9385 - f1_score: 0.1806 - val_loss: 1.3631 - val_categorical_accuracy: 0.5946 - val_precision_24: 0.1314 - val_recall_24: 0.9357 - val_f1_score: 0.1698
Epoch 12/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3804 - categorical_accuracy: 0.5884 - precision_24: 0.1300 - recall_24: 0.9411 - f1_score: 0.1858 - val_loss: 1.3576 - val_categorical_accuracy: 0.5972 - val_precision_24: 0.1305 - val_recall_24: 0.9373 - val_f1_score: 0.1769
Epoch 13/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3691 - categorical_accuracy: 0.5962 - precision_24: 0.1292 - recall_24: 0.9418 - f1_score: 0.1874 - val_loss: 1.3499 - val_categorical_accuracy: 0.5984 - val_precision_24: 0.1315 - val_recall_24: 0.9402 - val_f1_score: 0.1818
Epoch 14/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3603 - categorical_accuracy: 0.5949 - precision_24: 0.1298 - recall_24: 0.9457 - f1_score: 0.1890 - val_loss: 1.3394 - val_categorical_accuracy: 0.6009 - val_precision_24: 0.1312 - val_recall_24: 0.9418 - val_f1_score: 0.1828
Epoch 15/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3531 - categorical_accuracy: 0.5949 - precision_24: 0.1296 - recall_24: 0.9475 - f1_score: 0.1954 - val_loss: 1.3357 - val_categorical_accuracy: 0.6035 - val_precision_24: 0.1310 - val_recall_24: 0.9431 - val_f1_score: 0.1821
Epoch 16/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3472 - categorical_accuracy: 0.5995 - precision_24: 0.1293 - recall_24: 0.9482 - f1_score: 0.1986 - val_loss: 1.3283 - val_categorical_accuracy: 0.6056 - val_precision_24: 0.1303 - val_recall_24: 0.9433 - val_f1_score: 0.1813
Epoch 17/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.3404 - categorical_accuracy: 0.5990 - precision_24: 0.1295 - recall_24: 0.9504 - f1_score: 0.1982 - val_loss: 1.3309 - val_categorical_accuracy: 0.6076 - val_precision_24: 0.1296 - val_recall_24: 0.9437 - val_f1_score: 0.1804

 0 64 15 0.001 189 48 

Epoch 1/50

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

1367/1367 [==============================] - 19s 13ms/step - loss: 1.7370 - categorical_accuracy: 0.4994 - precision_25: 0.1245 - recall_25: 0.8601 - f1_score: 0.1576 - val_loss: 1.6419 - val_categorical_accuracy: 0.5438 - val_precision_25: 0.1282 - val_recall_25: 0.9136 - val_f1_score: 0.1645
Epoch 2/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.4032 - categorical_accuracy: 0.5844 - precision_25: 0.1306 - recall_25: 0.9371 - f1_score: 0.1783 - val_loss: 1.6448 - val_categorical_accuracy: 0.5102 - val_precision_25: 0.1312 - val_recall_25: 0.9236 - val_f1_score: 0.1712
Epoch 3/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.3638 - categorical_accuracy: 0.5918 - precision_25: 0.1311 - recall_25: 0.9454 - f1_score: 0.1878 - val_loss: 1.6282 - val_categorical_accuracy: 0.5192 - val_precision_25: 0.1305 - val_recall_25: 0.9319 - val_f1_score: 0.1721
Epoch 4/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.3382 - categorical_accuracy: 0.6092 - precision_25: 0.1311 - recall_25: 0.9525 - f1_score: 0.1946 - val_loss: 1.6145 - val_categorical_accuracy: 0.5511 - val_precision_25: 0.1256 - val_recall_25: 0.9322 - val_f1_score: 0.1787
Epoch 5/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.3254 - categorical_accuracy: 0.6049 - precision_25: 0.1305 - recall_25: 0.9548 - f1_score: 0.1987 - val_loss: 1.5625 - val_categorical_accuracy: 0.5893 - val_precision_25: 0.1271 - val_recall_25: 0.9387 - val_f1_score: 0.1806
Epoch 6/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.3277 - categorical_accuracy: 0.6144 - precision_25: 0.1297 - recall_25: 0.9532 - f1_score: 0.1998 - val_loss: 1.6057 - val_categorical_accuracy: 0.5736 - val_precision_25: 0.1240 - val_recall_25: 0.9336 - val_f1_score: 0.1757
Epoch 7/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.3060 - categorical_accuracy: 0.6233 - precision_25: 0.1299 - recall_25: 0.9588 - f1_score: 0.2044 - val_loss: 1.5539 - val_categorical_accuracy: 0.5837 - val_precision_25: 0.1246 - val_recall_25: 0.9399 - val_f1_score: 0.1795
Epoch 8/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.3033 - categorical_accuracy: 0.6301 - precision_25: 0.1294 - recall_25: 0.9550 - f1_score: 0.2028 - val_loss: 1.5692 - val_categorical_accuracy: 0.5831 - val_precision_25: 0.1260 - val_recall_25: 0.9348 - val_f1_score: 0.1848
Epoch 9/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.2952 - categorical_accuracy: 0.6233 - precision_25: 0.1309 - recall_25: 0.9572 - f1_score: 0.2122 - val_loss: 1.6200 - val_categorical_accuracy: 0.5490 - val_precision_25: 0.1288 - val_recall_25: 0.9454 - val_f1_score: 0.1876
Epoch 10/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.2904 - categorical_accuracy: 0.6218 - precision_25: 0.1311 - recall_25: 0.9555 - f1_score: 0.2119 - val_loss: 1.6166 - val_categorical_accuracy: 0.5655 - val_precision_25: 0.1282 - val_recall_25: 0.9405 - val_f1_score: 0.1822
Epoch 11/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.2874 - categorical_accuracy: 0.6267 - precision_25: 0.1318 - recall_25: 0.9551 - f1_score: 0.2151 - val_loss: 1.5793 - val_categorical_accuracy: 0.5738 - val_precision_25: 0.1269 - val_recall_25: 0.9378 - val_f1_score: 0.1783
Epoch 12/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.2906 - categorical_accuracy: 0.6262 - precision_25: 0.1324 - recall_25: 0.9516 - f1_score: 0.2161 - val_loss: 1.5970 - val_categorical_accuracy: 0.5811 - val_precision_25: 0.1275 - val_recall_25: 0.9465 - val_f1_score: 0.1849

 1 64 15 0.001 189 48 

Epoch 1/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.8104 - categorical_accuracy: 0.4827 - precision_26: 0.1229 - recall_26: 0.8523 - f1_score: 0.1526 - val_loss: 1.4663 - val_categorical_accuracy: 0.5501 - val_precision_26: 0.1267 - val_recall_26: 0.9200 - val_f1_score: 0.1663
Epoch 2/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.4272 - categorical_accuracy: 0.5824 - precision_26: 0.1366 - recall_26: 0.9306 - f1_score: 0.1819 - val_loss: 1.4190 - val_categorical_accuracy: 0.5952 - val_precision_26: 0.1316 - val_recall_26: 0.9374 - val_f1_score: 0.1856
Epoch 3/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3986 - categorical_accuracy: 0.5838 - precision_26: 0.1320 - recall_26: 0.9393 - f1_score: 0.2014 - val_loss: 1.3773 - val_categorical_accuracy: 0.5909 - val_precision_26: 0.1346 - val_recall_26: 0.9434 - val_f1_score: 0.1950
Epoch 4/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3658 - categorical_accuracy: 0.5985 - precision_26: 0.1342 - recall_26: 0.9442 - f1_score: 0.2116 - val_loss: 1.3968 - val_categorical_accuracy: 0.6128 - val_precision_26: 0.1381 - val_recall_26: 0.9467 - val_f1_score: 0.2055
Epoch 5/50
1371/1371 [==============================] - 18s 13ms/step - loss: 1.3623 - categorical_accuracy: 0.5975 - precision_26: 0.1347 - recall_26: 0.9444 - f1_score: 0.2168 - val_loss: 1.3735 - val_categorical_accuracy: 0.5830 - val_precision_26: 0.1333 - val_recall_26: 0.9509 - val_f1_score: 0.1990
Epoch 6/50
1371/1371 [==============================] - 17s 13ms/step - loss: 1.3457 - categorical_accuracy: 0.6054 - precision_26: 0.1337 - recall_26: 0.9482 - f1_score: 0.2254 - val_loss: 1.3706 - val_categorical_accuracy: 0.6215 - val_precision_26: 0.1315 - val_recall_26: 0.9402 - val_f1_score: 0.1924
Epoch 7/50
1371/1371 [==============================] - 17s 13ms/step - loss: 1.3412 - categorical_accuracy: 0.6076 - precision_26: 0.1344 - recall_26: 0.9475 - f1_score: 0.2308 - val_loss: 1.3447 - val_categorical_accuracy: 0.6353 - val_precision_26: 0.1354 - val_recall_26: 0.9393 - val_f1_score: 0.2040

 2 64 15 0.001 190 47 

Epoch 1/50
1304/1304 [==============================] - 19s 14ms/step - loss: 1.7553 - categorical_accuracy: 0.4993 - precision_27: 0.1276 - recall_27: 0.8447 - f1_score: 0.1500 - val_loss: 2.0312 - val_categorical_accuracy: 0.4775 - val_precision_27: 0.1228 - val_recall_27: 0.8339 - val_f1_score: 0.1478
Epoch 2/50
1304/1304 [==============================] - 18s 14ms/step - loss: 1.4024 - categorical_accuracy: 0.5822 - precision_27: 0.1379 - recall_27: 0.9279 - f1_score: 0.1766 - val_loss: 2.1022 - val_categorical_accuracy: 0.4573 - val_precision_27: 0.1205 - val_recall_27: 0.8548 - val_f1_score: 0.1536
Epoch 3/50
1304/1304 [==============================] - 18s 14ms/step - loss: 1.3509 - categorical_accuracy: 0.5944 - precision_27: 0.1344 - recall_27: 0.9382 - f1_score: 0.1905 - val_loss: 2.0769 - val_categorical_accuracy: 0.4743 - val_precision_27: 0.1204 - val_recall_27: 0.8702 - val_f1_score: 0.1542
Epoch 4/50
1304/1304 [==============================] - 18s 14ms/step - loss: 1.3311 - categorical_accuracy: 0.6144 - precision_27: 0.1351 - recall_27: 0.9434 - f1_score: 0.2012 - val_loss: 1.9735 - val_categorical_accuracy: 0.4768 - val_precision_27: 0.1254 - val_recall_27: 0.9087 - val_f1_score: 0.1711
Epoch 5/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.3197 - categorical_accuracy: 0.6099 - precision_27: 0.1331 - recall_27: 0.9493 - f1_score: 0.2023 - val_loss: 2.1227 - val_categorical_accuracy: 0.4719 - val_precision_27: 0.1214 - val_recall_27: 0.8706 - val_f1_score: 0.1603
Epoch 6/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.3131 - categorical_accuracy: 0.6152 - precision_27: 0.1345 - recall_27: 0.9442 - f1_score: 0.2054 - val_loss: 2.0533 - val_categorical_accuracy: 0.4887 - val_precision_27: 0.1239 - val_recall_27: 0.9126 - val_f1_score: 0.1658
Epoch 7/50
1304/1304 [==============================] - 17s 13ms/step - loss: 1.2993 - categorical_accuracy: 0.6168 - precision_27: 0.1343 - recall_27: 0.9490 - f1_score: 0.2078 - val_loss: 2.1295 - val_categorical_accuracy: 0.4678 - val_precision_27: 0.1261 - val_recall_27: 0.8967 - val_f1_score: 0.1655

 3 64 15 0.001 190 47 

Epoch 1/50
1387/1387 [==============================] - 19s 13ms/step - loss: 1.7735 - categorical_accuracy: 0.4939 - precision_28: 0.1216 - recall_28: 0.8582 - f1_score: 0.1547 - val_loss: 1.4852 - val_categorical_accuracy: 0.5848 - val_precision_28: 0.1272 - val_recall_28: 0.9095 - val_f1_score: 0.1648
Epoch 2/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.4354 - categorical_accuracy: 0.5707 - precision_28: 0.1319 - recall_28: 0.9430 - f1_score: 0.1834 - val_loss: 1.4334 - val_categorical_accuracy: 0.5862 - val_precision_28: 0.1312 - val_recall_28: 0.9329 - val_f1_score: 0.1812
Epoch 3/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3804 - categorical_accuracy: 0.5930 - precision_28: 0.1324 - recall_28: 0.9491 - f1_score: 0.2017 - val_loss: 1.4276 - val_categorical_accuracy: 0.5913 - val_precision_28: 0.1305 - val_recall_28: 0.9374 - val_f1_score: 0.1918
Epoch 4/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3683 - categorical_accuracy: 0.5888 - precision_28: 0.1307 - recall_28: 0.9518 - f1_score: 0.1992 - val_loss: 1.3629 - val_categorical_accuracy: 0.6399 - val_precision_28: 0.1288 - val_recall_28: 0.9315 - val_f1_score: 0.1840
Epoch 5/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3355 - categorical_accuracy: 0.6097 - precision_28: 0.1314 - recall_28: 0.9559 - f1_score: 0.2043 - val_loss: 1.4246 - val_categorical_accuracy: 0.6016 - val_precision_28: 0.1300 - val_recall_28: 0.9404 - val_f1_score: 0.1994
Epoch 6/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3443 - categorical_accuracy: 0.5955 - precision_28: 0.1304 - recall_28: 0.9575 - f1_score: 0.2052 - val_loss: 1.4210 - val_categorical_accuracy: 0.5904 - val_precision_28: 0.1322 - val_recall_28: 0.9411 - val_f1_score: 0.1990
Epoch 7/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3356 - categorical_accuracy: 0.6071 - precision_28: 0.1302 - recall_28: 0.9553 - f1_score: 0.2079 - val_loss: 1.3235 - val_categorical_accuracy: 0.6253 - val_precision_28: 0.1329 - val_recall_28: 0.9398 - val_f1_score: 0.1975
Epoch 8/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3271 - categorical_accuracy: 0.6076 - precision_28: 0.1313 - recall_28: 0.9568 - f1_score: 0.2122 - val_loss: 1.3706 - val_categorical_accuracy: 0.5701 - val_precision_28: 0.1296 - val_recall_28: 0.9378 - val_f1_score: 0.1943

 4 64 15 0.001 190 47 

Epoch 1/50
1387/1387 [==============================] - 19s 13ms/step - loss: 1.8188 - categorical_accuracy: 0.4767 - precision_29: 0.1266 - recall_29: 0.8543 - f1_score: 0.1588 - val_loss: 1.4056 - val_categorical_accuracy: 0.5757 - val_precision_29: 0.1440 - val_recall_29: 0.9354 - val_f1_score: 0.1719
Epoch 2/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.4604 - categorical_accuracy: 0.5648 - precision_29: 0.1346 - recall_29: 0.9353 - f1_score: 0.1863 - val_loss: 1.3600 - val_categorical_accuracy: 0.6264 - val_precision_29: 0.1441 - val_recall_29: 0.9482 - val_f1_score: 0.1816
Epoch 3/50
1387/1387 [==============================] - 17s 12ms/step - loss: 1.4025 - categorical_accuracy: 0.5861 - precision_29: 0.1342 - recall_29: 0.9485 - f1_score: 0.1986 - val_loss: 1.3789 - val_categorical_accuracy: 0.6100 - val_precision_29: 0.1353 - val_recall_29: 0.9513 - val_f1_score: 0.1793
Epoch 4/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3926 - categorical_accuracy: 0.5894 - precision_29: 0.1327 - recall_29: 0.9515 - f1_score: 0.2117 - val_loss: 1.3101 - val_categorical_accuracy: 0.6504 - val_precision_29: 0.1346 - val_recall_29: 0.9550 - val_f1_score: 0.1777
Epoch 5/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3590 - categorical_accuracy: 0.6034 - precision_29: 0.1330 - recall_29: 0.9536 - f1_score: 0.2142 - val_loss: 1.3624 - val_categorical_accuracy: 0.6150 - val_precision_29: 0.1355 - val_recall_29: 0.9558 - val_f1_score: 0.1873
Epoch 6/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3550 - categorical_accuracy: 0.6040 - precision_29: 0.1328 - recall_29: 0.9565 - f1_score: 0.2169 - val_loss: 1.3777 - val_categorical_accuracy: 0.6085 - val_precision_29: 0.1364 - val_recall_29: 0.9580 - val_f1_score: 0.1819
Epoch 7/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3554 - categorical_accuracy: 0.6009 - precision_29: 0.1320 - recall_29: 0.9546 - f1_score: 0.2302 - val_loss: 1.3181 - val_categorical_accuracy: 0.6130 - val_precision_29: 0.1382 - val_recall_29: 0.9578 - val_f1_score: 0.1895
Epoch 8/50
1387/1387 [==============================] - 17s 13ms/step - loss: 1.3404 - categorical_accuracy: 0.6061 - precision_29: 0.1334 - recall_29: 0.9578 - f1_score: 0.2311 - val_loss: 1.3259 - val_categorical_accuracy: 0.6224 - val_precision_29: 0.1327 - val_recall_29: 0.9585 - val_f1_score: 0.1845
Epoch 9/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3254 - categorical_accuracy: 0.6097 - precision_29: 0.1323 - recall_29: 0.9591 - f1_score: 0.2321 - val_loss: 1.3274 - val_categorical_accuracy: 0.6257 - val_precision_29: 0.1371 - val_recall_29: 0.9536 - val_f1_score: 0.1906
Epoch 10/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3335 - categorical_accuracy: 0.6123 - precision_29: 0.1336 - recall_29: 0.9577 - f1_score: 0.2334 - val_loss: 1.2943 - val_categorical_accuracy: 0.6315 - val_precision_29: 0.1355 - val_recall_29: 0.9525 - val_f1_score: 0.1905
Epoch 11/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3252 - categorical_accuracy: 0.6133 - precision_29: 0.1343 - recall_29: 0.9536 - f1_score: 0.2329 - val_loss: 1.3046 - val_categorical_accuracy: 0.6182 - val_precision_29: 0.1363 - val_recall_29: 0.9631 - val_f1_score: 0.1876
Epoch 12/50
1387/1387 [==============================] - 18s 13ms/step - loss: 1.3214 - categorical_accuracy: 0.6054 - precision_29: 0.1327 - recall_29: 0.9549 - f1_score: 0.2471 - val_loss: 1.3375 - val_categorical_accuracy: 0.6324 - val_precision_29: 0.1335 - val_recall_29: 0.9539 - val_f1_score: 0.1858

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

 0 64 30 0.0001 189 48 

Epoch 1/50
1363/1363 [==============================] - 50s 36ms/step - loss: 2.3326 - categorical_accuracy: 0.3649 - precision_30: 0.0965 - recall_30: 0.7332 - f1_score: 0.1388 - val_loss: 2.0713 - val_categorical_accuracy: 0.4302 - val_precision_30: 0.1009 - val_recall_30: 0.8006 - val_f1_score: 0.1394
Epoch 2/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.7957 - categorical_accuracy: 0.4867 - precision_30: 0.1114 - recall_30: 0.8380 - f1_score: 0.1527 - val_loss: 1.8288 - val_categorical_accuracy: 0.4880 - val_precision_30: 0.1146 - val_recall_30: 0.8499 - val_f1_score: 0.1438
Epoch 3/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.5923 - categorical_accuracy: 0.5435 - precision_30: 0.1235 - recall_30: 0.8798 - f1_score: 0.1570 - val_loss: 1.7159 - val_categorical_accuracy: 0.5086 - val_precision_30: 0.1231 - val_recall_30: 0.8756 - val_f1_score: 0.1499
Epoch 4/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.4849 - categorical_accuracy: 0.5722 - precision_30: 0.1291 - recall_30: 0.9025 - f1_score: 0.1611 - val_loss: 1.6380 - val_categorical_accuracy: 0.5264 - val_precision_30: 0.1285 - val_recall_30: 0.8923 - val_f1_score: 0.1570
Epoch 5/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.4230 - categorical_accuracy: 0.5820 - precision_30: 0.1334 - recall_30: 0.9196 - f1_score: 0.1684 - val_loss: 1.6051 - val_categorical_accuracy: 0.5368 - val_precision_30: 0.1295 - val_recall_30: 0.9034 - val_f1_score: 0.1676
Epoch 6/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.3798 - categorical_accuracy: 0.5865 - precision_30: 0.1340 - recall_30: 0.9265 - f1_score: 0.1743 - val_loss: 1.5842 - val_categorical_accuracy: 0.5383 - val_precision_30: 0.1316 - val_recall_30: 0.9130 - val_f1_score: 0.1731
Epoch 7/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.3499 - categorical_accuracy: 0.5973 - precision_30: 0.1343 - recall_30: 0.9323 - f1_score: 0.1782 - val_loss: 1.5624 - val_categorical_accuracy: 0.5417 - val_precision_30: 0.1325 - val_recall_30: 0.9223 - val_f1_score: 0.1764
Epoch 8/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.3228 - categorical_accuracy: 0.6023 - precision_30: 0.1351 - recall_30: 0.9370 - f1_score: 0.1834 - val_loss: 1.5516 - val_categorical_accuracy: 0.5362 - val_precision_30: 0.1315 - val_recall_30: 0.9243 - val_f1_score: 0.1787
Epoch 9/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.3087 - categorical_accuracy: 0.6113 - precision_30: 0.1344 - recall_30: 0.9405 - f1_score: 0.1842 - val_loss: 1.5447 - val_categorical_accuracy: 0.5377 - val_precision_30: 0.1321 - val_recall_30: 0.9283 - val_f1_score: 0.1801
Epoch 10/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2950 - categorical_accuracy: 0.6088 - precision_30: 0.1337 - recall_30: 0.9436 - f1_score: 0.1831 - val_loss: 1.5445 - val_categorical_accuracy: 0.5354 - val_precision_30: 0.1323 - val_recall_30: 0.9310 - val_f1_score: 0.1858
Epoch 11/50
1363/1363 [==============================] - 20s 14ms/step - loss: 1.2837 - categorical_accuracy: 0.6181 - precision_30: 0.1334 - recall_30: 0.9451 - f1_score: 0.1867 - val_loss: 1.5393 - val_categorical_accuracy: 0.5412 - val_precision_30: 0.1314 - val_recall_30: 0.9337 - val_f1_score: 0.1851
Epoch 12/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2719 - categorical_accuracy: 0.6146 - precision_30: 0.1343 - recall_30: 0.9478 - f1_score: 0.1885 - val_loss: 1.5393 - val_categorical_accuracy: 0.5400 - val_precision_30: 0.1310 - val_recall_30: 0.9330 - val_f1_score: 0.1862
Epoch 13/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2655 - categorical_accuracy: 0.6195 - precision_30: 0.1336 - recall_30: 0.9475 - f1_score: 0.1903 - val_loss: 1.5294 - val_categorical_accuracy: 0.5488 - val_precision_30: 0.1311 - val_recall_30: 0.9363 - val_f1_score: 0.1867
Epoch 14/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2559 - categorical_accuracy: 0.6192 - precision_30: 0.1325 - recall_30: 0.9483 - f1_score: 0.1890 - val_loss: 1.5285 - val_categorical_accuracy: 0.5452 - val_precision_30: 0.1306 - val_recall_30: 0.9376 - val_f1_score: 0.1877
Epoch 15/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2488 - categorical_accuracy: 0.6213 - precision_30: 0.1328 - recall_30: 0.9510 - f1_score: 0.1930 - val_loss: 1.5496 - val_categorical_accuracy: 0.5260 - val_precision_30: 0.1320 - val_recall_30: 0.9392 - val_f1_score: 0.1892
Epoch 16/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2458 - categorical_accuracy: 0.6237 - precision_30: 0.1319 - recall_30: 0.9511 - f1_score: 0.1919 - val_loss: 1.5349 - val_categorical_accuracy: 0.5513 - val_precision_30: 0.1302 - val_recall_30: 0.9364 - val_f1_score: 0.1890
Epoch 17/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2389 - categorical_accuracy: 0.6276 - precision_30: 0.1330 - recall_30: 0.9517 - f1_score: 0.1952 - val_loss: 1.5390 - val_categorical_accuracy: 0.5534 - val_precision_30: 0.1286 - val_recall_30: 0.9387 - val_f1_score: 0.1877
Epoch 18/50
1363/1363 [==============================] - 20s 15ms/step - loss: 1.2346 - categorical_accuracy: 0.6252 - precision_30: 0.1323 - recall_30: 0.9520 - f1_score: 0.1953 - val_loss: 1.5447 - val_categorical_accuracy: 0.5512 - val_precision_30: 0.1285 - val_recall_30: 0.9394 - val_f1_score: 0.1876

 1 64 30 0.0001 189 48 

Epoch 1/50
1367/1367 [==============================] - 21s 15ms/step - loss: 2.3317 - categorical_accuracy: 0.3581 - precision_31: 0.0971 - recall_31: 0.7458 - f1_score: 0.1383 - val_loss: 1.9679 - val_categorical_accuracy: 0.4611 - val_precision_31: 0.1078 - val_recall_31: 0.8001 - val_f1_score: 0.1445
Epoch 2/50
1367/1367 [==============================] - 20s 15ms/step - loss: 1.8171 - categorical_accuracy: 0.4923 - precision_31: 0.1136 - recall_31: 0.8355 - f1_score: 0.1513 - val_loss: 1.7253 - val_categorical_accuracy: 0.5186 - val_precision_31: 0.1174 - val_recall_31: 0.8437 - val_f1_score: 0.1503
Epoch 3/50
1367/1367 [==============================] - 20s 15ms/step - loss: 1.6265 - categorical_accuracy: 0.5362 - precision_31: 0.1214 - recall_31: 0.8782 - f1_score: 0.1562 - val_loss: 1.6051 - val_categorical_accuracy: 0.5418 - val_precision_31: 0.1208 - val_recall_31: 0.8717 - val_f1_score: 0.1544
Epoch 4/50
1367/1367 [==============================] - 20s 15ms/step - loss: 1.5109 - categorical_accuracy: 0.5596 - precision_31: 0.1266 - recall_31: 0.9056 - f1_score: 0.1611 - val_loss: 1.5305 - val_categorical_accuracy: 0.5557 - val_precision_31: 0.1248 - val_recall_31: 0.8909 - val_f1_score: 0.1599
Epoch 5/50
1367/1367 [==============================] - 20s 15ms/step - loss: 1.4461 - categorical_accuracy: 0.5703 - precision_31: 0.1298 - recall_31: 0.9191 - f1_score: 0.1663 - val_loss: 1.4833 - val_categorical_accuracy: 0.5628 - val_precision_31: 0.1276 - val_recall_31: 0.9110 - val_f1_score: 0.1667
Epoch 6/50
1367/1367 [==============================] - 20s 15ms/step - loss: 1.4046 - categorical_accuracy: 0.5804 - precision_31: 0.1301 - recall_31: 0.9267 - f1_score: 0.1700 - val_loss: 1.4595 - val_categorical_accuracy: 0.5572 - val_precision_31: 0.1298 - val_recall_31: 0.9192 - val_f1_score: 0.1726
Epoch 7/50
1367/1367 [==============================] - 31s 23ms/step - loss: 1.3722 - categorical_accuracy: 0.5843 - precision_31: 0.1317 - recall_31: 0.9336 - f1_score: 0.1755 - val_loss: 1.4310 - val_categorical_accuracy: 0.5602 - val_precision_31: 0.1298 - val_recall_31: 0.9254 - val_f1_score: 0.1748
Epoch 8/50
1367/1367 [==============================] - 19s 14ms/step - loss: 1.3502 - categorical_accuracy: 0.5916 - precision_31: 0.1314 - recall_31: 0.9365 - f1_score: 0.1790 - val_loss: 1.4140 - val_categorical_accuracy: 0.5662 - val_precision_31: 0.1310 - val_recall_31: 0.9300 - val_f1_score: 0.1784
Epoch 9/50
1367/1367 [==============================] - 17s 12ms/step - loss: 1.3334 - categorical_accuracy: 0.5944 - precision_31: 0.1323 - recall_31: 0.9385 - f1_score: 0.1815 - val_loss: 1.3888 - val_categorical_accuracy: 0.5843 - val_precision_31: 0.1307 - val_recall_31: 0.9344 - val_f1_score: 0.1798
Epoch 10/50
1367/1367 [==============================] - 17s 12ms/step - loss: 1.3173 - categorical_accuracy: 0.5987 - precision_31: 0.1322 - recall_31: 0.9403 - f1_score: 0.1858 - val_loss: 1.3790 - val_categorical_accuracy: 0.5809 - val_precision_31: 0.1306 - val_recall_31: 0.9385 - val_f1_score: 0.1815
Epoch 11/50
1367/1367 [==============================] - 17s 12ms/step - loss: 1.3057 - categorical_accuracy: 0.6014 - precision_31: 0.1321 - recall_31: 0.9433 - f1_score: 0.1859 - val_loss: 1.3644 - val_categorical_accuracy: 0.5740 - val_precision_31: 0.1314 - val_recall_31: 0.9427 - val_f1_score: 0.1838
Epoch 12/50
1367/1367 [==============================] - 17s 13ms/step - loss: 1.2942 - categorical_accuracy: 0.6017 - precision_31: 0.1321 - recall_31: 0.9436 - f1_score: 0.1896 - val_loss: 1.3543 - val_categorical_accuracy: 0.5813 - val_precision_31: 0.1318 - val_recall_31: 0.9438 - val_f1_score: 0.1840
Epoch 13/50
1367/1367 [==============================] - 17s 12ms/step - loss: 1.2841 - categorical_accuracy: 0.6099 - precision_31: 0.1317 - recall_31: 0.9456 - f1_score: 0.1930 - val_loss: 1.3560 - val_categorical_accuracy: 0.5858 - val_precision_31: 0.1311 - val_recall_31: 0.9458 - val_f1_score: 0.1844
Epoch 14/50
1367/1367 [==============================] - 18s 13ms/step - loss: 1.2773 - categorical_accuracy: 0.6110 - precision_31: 0.1314 - recall_31: 0.9464 - f1_score: 0.1945 - val_loss: 1.3484 - val_categorical_accuracy: 0.5812 - val_precision_31: 0.1316 - val_recall_31: 0.9495 - val_f1_score: 0.1868
Epoch 15/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2729 - categorical_accuracy: 0.6137 - precision_31: 0.1320 - recall_31: 0.9472 - f1_score: 0.2008 - val_loss: 1.3371 - val_categorical_accuracy: 0.5833 - val_precision_31: 0.1310 - val_recall_31: 0.9523 - val_f1_score: 0.1879
Epoch 16/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2639 - categorical_accuracy: 0.6165 - precision_31: 0.1315 - recall_31: 0.9478 - f1_score: 0.1995 - val_loss: 1.3343 - val_categorical_accuracy: 0.5917 - val_precision_31: 0.1306 - val_recall_31: 0.9527 - val_f1_score: 0.1872
Epoch 17/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.2586 - categorical_accuracy: 0.6167 - precision_31: 0.1310 - recall_31: 0.9490 - f1_score: 0.2053 - val_loss: 1.3323 - val_categorical_accuracy: 0.5870 - val_precision_31: 0.1314 - val_recall_31: 0.9528 - val_f1_score: 0.1895
Epoch 18/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2573 - categorical_accuracy: 0.6162 - precision_31: 0.1307 - recall_31: 0.9508 - f1_score: 0.2051 - val_loss: 1.3213 - val_categorical_accuracy: 0.5933 - val_precision_31: 0.1313 - val_recall_31: 0.9533 - val_f1_score: 0.1895
Epoch 19/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.2510 - categorical_accuracy: 0.6167 - precision_31: 0.1317 - recall_31: 0.9500 - f1_score: 0.2071 - val_loss: 1.3134 - val_categorical_accuracy: 0.5934 - val_precision_31: 0.1310 - val_recall_31: 0.9564 - val_f1_score: 0.1886
Epoch 20/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2469 - categorical_accuracy: 0.6188 - precision_31: 0.1310 - recall_31: 0.9522 - f1_score: 0.2129 - val_loss: 1.3135 - val_categorical_accuracy: 0.5982 - val_precision_31: 0.1302 - val_recall_31: 0.9543 - val_f1_score: 0.1884
Epoch 21/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2399 - categorical_accuracy: 0.6258 - precision_31: 0.1305 - recall_31: 0.9514 - f1_score: 0.2117 - val_loss: 1.3117 - val_categorical_accuracy: 0.5979 - val_precision_31: 0.1312 - val_recall_31: 0.9528 - val_f1_score: 0.1903
Epoch 22/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2409 - categorical_accuracy: 0.6280 - precision_31: 0.1312 - recall_31: 0.9525 - f1_score: 0.2139 - val_loss: 1.3117 - val_categorical_accuracy: 0.5940 - val_precision_31: 0.1305 - val_recall_31: 0.9550 - val_f1_score: 0.1905
Epoch 23/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2379 - categorical_accuracy: 0.6200 - precision_31: 0.1306 - recall_31: 0.9527 - f1_score: 0.2152 - val_loss: 1.3142 - val_categorical_accuracy: 0.5913 - val_precision_31: 0.1312 - val_recall_31: 0.9552 - val_f1_score: 0.1909
Epoch 24/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2343 - categorical_accuracy: 0.6280 - precision_31: 0.1304 - recall_31: 0.9528 - f1_score: 0.2166 - val_loss: 1.3052 - val_categorical_accuracy: 0.5944 - val_precision_31: 0.1313 - val_recall_31: 0.9556 - val_f1_score: 0.1914
Epoch 25/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2307 - categorical_accuracy: 0.6273 - precision_31: 0.1307 - recall_31: 0.9536 - f1_score: 0.2193 - val_loss: 1.3063 - val_categorical_accuracy: 0.5920 - val_precision_31: 0.1323 - val_recall_31: 0.9558 - val_f1_score: 0.1930
Epoch 26/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2307 - categorical_accuracy: 0.6286 - precision_31: 0.1312 - recall_31: 0.9526 - f1_score: 0.2227 - val_loss: 1.3106 - val_categorical_accuracy: 0.5964 - val_precision_31: 0.1311 - val_recall_31: 0.9567 - val_f1_score: 0.1939
Epoch 27/50
1367/1367 [==============================] - 13s 10ms/step - loss: 1.2260 - categorical_accuracy: 0.6305 - precision_31: 0.1308 - recall_31: 0.9527 - f1_score: 0.2211 - val_loss: 1.3113 - val_categorical_accuracy: 0.5963 - val_precision_31: 0.1306 - val_recall_31: 0.9603 - val_f1_score: 0.1910
Epoch 28/50
1367/1367 [==============================] - 14s 10ms/step - loss: 1.2248 - categorical_accuracy: 0.6289 - precision_31: 0.1302 - recall_31: 0.9549 - f1_score: 0.2274 - val_loss: 1.2969 - val_categorical_accuracy: 0.5995 - val_precision_31: 0.1306 - val_recall_31: 0.9595 - val_f1_score: 0.1922
Epoch 29/50
1367/1367 [==============================] - 14s 10ms/step - loss: 1.2199 - categorical_accuracy: 0.6354 - precision_31: 0.1303 - recall_31: 0.9537 - f1_score: 0.2242 - val_loss: 1.2983 - val_categorical_accuracy: 0.5967 - val_precision_31: 0.1309 - val_recall_31: 0.9601 - val_f1_score: 0.1930

 2 64 30 0.0001 190 47 

Epoch 1/50
1302/1302 [==============================] - 18s 13ms/step - loss: 2.4209 - categorical_accuracy: 0.3506 - precision_32: 0.0975 - recall_32: 0.7196 - f1_score: 0.1352 - val_loss: 2.2467 - val_categorical_accuracy: 0.3611 - val_precision_32: 0.1020 - val_recall_32: 0.7515 - val_f1_score: 0.1417
Epoch 2/50
1302/1302 [==============================] - 16s 13ms/step - loss: 1.8112 - categorical_accuracy: 0.4912 - precision_32: 0.1159 - recall_32: 0.8239 - f1_score: 0.1463 - val_loss: 2.0756 - val_categorical_accuracy: 0.3906 - val_precision_32: 0.1080 - val_recall_32: 0.7863 - val_f1_score: 0.1400
Epoch 3/50
1302/1302 [==============================] - 16s 13ms/step - loss: 1.5951 - categorical_accuracy: 0.5373 - precision_32: 0.1260 - recall_32: 0.8830 - f1_score: 0.1523 - val_loss: 2.0111 - val_categorical_accuracy: 0.4413 - val_precision_32: 0.1143 - val_recall_32: 0.8176 - val_f1_score: 0.1417
Epoch 4/50
1302/1302 [==============================] - 17s 13ms/step - loss: 1.4800 - categorical_accuracy: 0.5692 - precision_32: 0.1318 - recall_32: 0.9100 - f1_score: 0.1586 - val_loss: 2.0010 - val_categorical_accuracy: 0.4370 - val_precision_32: 0.1184 - val_recall_32: 0.8329 - val_f1_score: 0.1450
Epoch 5/50
1302/1302 [==============================] - 17s 13ms/step - loss: 1.4127 - categorical_accuracy: 0.5730 - precision_32: 0.1324 - recall_32: 0.9245 - f1_score: 0.1626 - val_loss: 2.0186 - val_categorical_accuracy: 0.4429 - val_precision_32: 0.1213 - val_recall_32: 0.8506 - val_f1_score: 0.1492
Epoch 6/50
1302/1302 [==============================] - 17s 13ms/step - loss: 1.3680 - categorical_accuracy: 0.5803 - precision_32: 0.1323 - recall_32: 0.9320 - f1_score: 0.1669 - val_loss: 2.0137 - val_categorical_accuracy: 0.4516 - val_precision_32: 0.1199 - val_recall_32: 0.8580 - val_f1_score: 0.1480
Epoch 7/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.3399 - categorical_accuracy: 0.5843 - precision_32: 0.1325 - recall_32: 0.9379 - f1_score: 0.1715 - val_loss: 2.0300 - val_categorical_accuracy: 0.4485 - val_precision_32: 0.1181 - val_recall_32: 0.8643 - val_f1_score: 0.1468
Epoch 8/50
1302/1302 [==============================] - 17s 13ms/step - loss: 1.3133 - categorical_accuracy: 0.5942 - precision_32: 0.1310 - recall_32: 0.9428 - f1_score: 0.1744 - val_loss: 2.0640 - val_categorical_accuracy: 0.4451 - val_precision_32: 0.1185 - val_recall_32: 0.8731 - val_f1_score: 0.1477

 3 64 30 0.0001 190 47 

Epoch 1/50
1384/1384 [==============================] - 19s 13ms/step - loss: 2.4260 - categorical_accuracy: 0.3423 - precision_33: 0.0961 - recall_33: 0.7239 - f1_score: 0.1364 - val_loss: 1.9432 - val_categorical_accuracy: 0.4649 - val_precision_33: 0.1142 - val_recall_33: 0.8168 - val_f1_score: 0.1479
Epoch 2/50
1384/1384 [==============================] - 17s 12ms/step - loss: 1.8568 - categorical_accuracy: 0.4728 - precision_33: 0.1167 - recall_33: 0.8387 - f1_score: 0.1550 - val_loss: 1.7105 - val_categorical_accuracy: 0.5213 - val_precision_33: 0.1185 - val_recall_33: 0.8561 - val_f1_score: 0.1485
Epoch 3/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.6528 - categorical_accuracy: 0.5282 - precision_33: 0.1239 - recall_33: 0.8739 - f1_score: 0.1598 - val_loss: 1.5699 - val_categorical_accuracy: 0.5497 - val_precision_33: 0.1238 - val_recall_33: 0.8819 - val_f1_score: 0.1526
Epoch 4/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.5392 - categorical_accuracy: 0.5516 - precision_33: 0.1286 - recall_33: 0.8988 - f1_score: 0.1651 - val_loss: 1.4888 - val_categorical_accuracy: 0.5631 - val_precision_33: 0.1269 - val_recall_33: 0.9024 - val_f1_score: 0.1585
Epoch 5/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.4663 - categorical_accuracy: 0.5638 - precision_33: 0.1303 - recall_33: 0.9143 - f1_score: 0.1692 - val_loss: 1.4410 - val_categorical_accuracy: 0.5802 - val_precision_33: 0.1303 - val_recall_33: 0.9131 - val_f1_score: 0.1646
Epoch 6/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.4208 - categorical_accuracy: 0.5739 - precision_33: 0.1317 - recall_33: 0.9276 - f1_score: 0.1759 - val_loss: 1.4095 - val_categorical_accuracy: 0.5823 - val_precision_33: 0.1284 - val_recall_33: 0.9214 - val_f1_score: 0.1696
Epoch 7/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3822 - categorical_accuracy: 0.5834 - precision_33: 0.1312 - recall_33: 0.9360 - f1_score: 0.1791 - val_loss: 1.3941 - val_categorical_accuracy: 0.5860 - val_precision_33: 0.1316 - val_recall_33: 0.9263 - val_f1_score: 0.1733
Epoch 8/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3575 - categorical_accuracy: 0.5898 - precision_33: 0.1328 - recall_33: 0.9418 - f1_score: 0.1849 - val_loss: 1.3920 - val_categorical_accuracy: 0.5715 - val_precision_33: 0.1305 - val_recall_33: 0.9296 - val_f1_score: 0.1769
Epoch 9/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3401 - categorical_accuracy: 0.5914 - precision_33: 0.1325 - recall_33: 0.9448 - f1_score: 0.1893 - val_loss: 1.3730 - val_categorical_accuracy: 0.5809 - val_precision_33: 0.1305 - val_recall_33: 0.9332 - val_f1_score: 0.1784
Epoch 10/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3201 - categorical_accuracy: 0.5928 - precision_33: 0.1323 - recall_33: 0.9515 - f1_score: 0.1905 - val_loss: 1.3674 - val_categorical_accuracy: 0.5860 - val_precision_33: 0.1312 - val_recall_33: 0.9347 - val_f1_score: 0.1815
Epoch 11/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3087 - categorical_accuracy: 0.5954 - precision_33: 0.1326 - recall_33: 0.9528 - f1_score: 0.1955 - val_loss: 1.3527 - val_categorical_accuracy: 0.5924 - val_precision_33: 0.1302 - val_recall_33: 0.9351 - val_f1_score: 0.1805
Epoch 12/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.2966 - categorical_accuracy: 0.6019 - precision_33: 0.1322 - recall_33: 0.9538 - f1_score: 0.1992 - val_loss: 1.3450 - val_categorical_accuracy: 0.5932 - val_precision_33: 0.1313 - val_recall_33: 0.9366 - val_f1_score: 0.1828
Epoch 13/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.2884 - categorical_accuracy: 0.6025 - precision_33: 0.1321 - recall_33: 0.9564 - f1_score: 0.2043 - val_loss: 1.3322 - val_categorical_accuracy: 0.5996 - val_precision_33: 0.1302 - val_recall_33: 0.9368 - val_f1_score: 0.1830
Epoch 14/50
1384/1384 [==============================] - 17s 12ms/step - loss: 1.2768 - categorical_accuracy: 0.6037 - precision_33: 0.1327 - recall_33: 0.9615 - f1_score: 0.2041 - val_loss: 1.3303 - val_categorical_accuracy: 0.6021 - val_precision_33: 0.1306 - val_recall_33: 0.9377 - val_f1_score: 0.1848
Epoch 15/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.2680 - categorical_accuracy: 0.6080 - precision_33: 0.1325 - recall_33: 0.9571 - f1_score: 0.2113 - val_loss: 1.3314 - val_categorical_accuracy: 0.5901 - val_precision_33: 0.1316 - val_recall_33: 0.9391 - val_f1_score: 0.1886
Epoch 16/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2627 - categorical_accuracy: 0.6090 - precision_33: 0.1331 - recall_33: 0.9642 - f1_score: 0.2100 - val_loss: 1.3232 - val_categorical_accuracy: 0.6014 - val_precision_33: 0.1307 - val_recall_33: 0.9405 - val_f1_score: 0.1867
Epoch 17/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2599 - categorical_accuracy: 0.6119 - precision_33: 0.1326 - recall_33: 0.9622 - f1_score: 0.2085 - val_loss: 1.3228 - val_categorical_accuracy: 0.5968 - val_precision_33: 0.1303 - val_recall_33: 0.9400 - val_f1_score: 0.1861
Epoch 18/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.2538 - categorical_accuracy: 0.6120 - precision_33: 0.1325 - recall_33: 0.9614 - f1_score: 0.2199 - val_loss: 1.3275 - val_categorical_accuracy: 0.5966 - val_precision_33: 0.1308 - val_recall_33: 0.9418 - val_f1_score: 0.1903
Epoch 19/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.2461 - categorical_accuracy: 0.6175 - precision_33: 0.1330 - recall_33: 0.9641 - f1_score: 0.2174 - val_loss: 1.3261 - val_categorical_accuracy: 0.5967 - val_precision_33: 0.1315 - val_recall_33: 0.9423 - val_f1_score: 0.1923
Epoch 20/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2447 - categorical_accuracy: 0.6115 - precision_33: 0.1328 - recall_33: 0.9669 - f1_score: 0.2171 - val_loss: 1.3111 - val_categorical_accuracy: 0.6164 - val_precision_33: 0.1309 - val_recall_33: 0.9404 - val_f1_score: 0.1905
Epoch 21/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2404 - categorical_accuracy: 0.6198 - precision_33: 0.1332 - recall_33: 0.9656 - f1_score: 0.2218 - val_loss: 1.3091 - val_categorical_accuracy: 0.6056 - val_precision_33: 0.1306 - val_recall_33: 0.9415 - val_f1_score: 0.1903
Epoch 22/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.2367 - categorical_accuracy: 0.6176 - precision_33: 0.1327 - recall_33: 0.9677 - f1_score: 0.2222 - val_loss: 1.3068 - val_categorical_accuracy: 0.6101 - val_precision_33: 0.1307 - val_recall_33: 0.9408 - val_f1_score: 0.1919

 4 64 30 0.0001 190 47 

Epoch 1/50
1384/1384 [==============================] - 18s 12ms/step - loss: 2.4734 - categorical_accuracy: 0.3306 - precision_34: 0.0956 - recall_34: 0.7151 - f1_score: 0.1358 - val_loss: 1.9165 - val_categorical_accuracy: 0.4661 - val_precision_34: 0.1147 - val_recall_34: 0.8133 - val_f1_score: 0.1434
Epoch 2/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.8413 - categorical_accuracy: 0.4699 - precision_34: 0.1168 - recall_34: 0.8482 - f1_score: 0.1568 - val_loss: 1.6380 - val_categorical_accuracy: 0.5421 - val_precision_34: 0.1280 - val_recall_34: 0.8905 - val_f1_score: 0.1544
Epoch 3/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.6363 - categorical_accuracy: 0.5315 - precision_34: 0.1258 - recall_34: 0.8884 - f1_score: 0.1618 - val_loss: 1.5029 - val_categorical_accuracy: 0.5565 - val_precision_34: 0.1350 - val_recall_34: 0.9154 - val_f1_score: 0.1609
Epoch 4/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.5185 - categorical_accuracy: 0.5514 - precision_34: 0.1320 - recall_34: 0.9121 - f1_score: 0.1672 - val_loss: 1.4254 - val_categorical_accuracy: 0.5802 - val_precision_34: 0.1374 - val_recall_34: 0.9280 - val_f1_score: 0.1687
Epoch 5/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.4551 - categorical_accuracy: 0.5689 - precision_34: 0.1343 - recall_34: 0.9250 - f1_score: 0.1743 - val_loss: 1.3807 - val_categorical_accuracy: 0.5853 - val_precision_34: 0.1341 - val_recall_34: 0.9352 - val_f1_score: 0.1664
Epoch 6/50
1384/1384 [==============================] - 26s 19ms/step - loss: 1.4087 - categorical_accuracy: 0.5730 - precision_34: 0.1327 - recall_34: 0.9325 - f1_score: 0.1752 - val_loss: 1.3640 - val_categorical_accuracy: 0.5931 - val_precision_34: 0.1351 - val_recall_34: 0.9409 - val_f1_score: 0.1724
Epoch 7/50
1384/1384 [==============================] - 18s 13ms/step - loss: 1.3772 - categorical_accuracy: 0.5828 - precision_34: 0.1329 - recall_34: 0.9385 - f1_score: 0.1810 - val_loss: 1.3319 - val_categorical_accuracy: 0.5958 - val_precision_34: 0.1345 - val_recall_34: 0.9441 - val_f1_score: 0.1743
Epoch 8/50
1384/1384 [==============================] - 14s 10ms/step - loss: 1.3566 - categorical_accuracy: 0.5842 - precision_34: 0.1332 - recall_34: 0.9398 - f1_score: 0.1851 - val_loss: 1.3160 - val_categorical_accuracy: 0.6076 - val_precision_34: 0.1344 - val_recall_34: 0.9466 - val_f1_score: 0.1746
Epoch 9/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3366 - categorical_accuracy: 0.5882 - precision_34: 0.1333 - recall_34: 0.9451 - f1_score: 0.1859 - val_loss: 1.3121 - val_categorical_accuracy: 0.5994 - val_precision_34: 0.1345 - val_recall_34: 0.9479 - val_f1_score: 0.1744
Epoch 10/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.3195 - categorical_accuracy: 0.5991 - precision_34: 0.1331 - recall_34: 0.9443 - f1_score: 0.1874 - val_loss: 1.2975 - val_categorical_accuracy: 0.6120 - val_precision_34: 0.1377 - val_recall_34: 0.9492 - val_f1_score: 0.1769
Epoch 11/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.3131 - categorical_accuracy: 0.5941 - precision_34: 0.1340 - recall_34: 0.9484 - f1_score: 0.1928 - val_loss: 1.2851 - val_categorical_accuracy: 0.6148 - val_precision_34: 0.1346 - val_recall_34: 0.9504 - val_f1_score: 0.1773
Epoch 12/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.3015 - categorical_accuracy: 0.6028 - precision_34: 0.1334 - recall_34: 0.9486 - f1_score: 0.1941 - val_loss: 1.2754 - val_categorical_accuracy: 0.6203 - val_precision_34: 0.1352 - val_recall_34: 0.9512 - val_f1_score: 0.1777
Epoch 13/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2951 - categorical_accuracy: 0.5959 - precision_34: 0.1337 - recall_34: 0.9507 - f1_score: 0.1992 - val_loss: 1.2808 - val_categorical_accuracy: 0.6139 - val_precision_34: 0.1343 - val_recall_34: 0.9522 - val_f1_score: 0.1797
Epoch 14/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.2849 - categorical_accuracy: 0.6044 - precision_34: 0.1329 - recall_34: 0.9507 - f1_score: 0.1969 - val_loss: 1.2664 - val_categorical_accuracy: 0.6219 - val_precision_34: 0.1349 - val_recall_34: 0.9523 - val_f1_score: 0.1814
Epoch 15/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.2821 - categorical_accuracy: 0.6047 - precision_34: 0.1333 - recall_34: 0.9537 - f1_score: 0.2026 - val_loss: 1.2685 - val_categorical_accuracy: 0.6245 - val_precision_34: 0.1345 - val_recall_34: 0.9527 - val_f1_score: 0.1781
Epoch 16/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.2754 - categorical_accuracy: 0.6070 - precision_34: 0.1331 - recall_34: 0.9540 - f1_score: 0.2051 - val_loss: 1.2576 - val_categorical_accuracy: 0.6214 - val_precision_34: 0.1347 - val_recall_34: 0.9533 - val_f1_score: 0.1786
Epoch 17/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2677 - categorical_accuracy: 0.6115 - precision_34: 0.1336 - recall_34: 0.9518 - f1_score: 0.2061 - val_loss: 1.2510 - val_categorical_accuracy: 0.6269 - val_precision_34: 0.1335 - val_recall_34: 0.9541 - val_f1_score: 0.1791

 0 64 30 0.001 189 48 

Epoch 1/50

C:\ProgramData\Anaconda3\lib\site-packages\sklearn\model_selection\_split.py:684: UserWarning: The least populated class in y has only 2 members, which is less than n_splits=5.
  warnings.warn(

1363/1363 [==============================] - 18s 12ms/step - loss: 1.7114 - categorical_accuracy: 0.5182 - precision_35: 0.1278 - recall_35: 0.8797 - f1_score: 0.1621 - val_loss: 1.6433 - val_categorical_accuracy: 0.5247 - val_precision_35: 0.1319 - val_recall_35: 0.9209 - val_f1_score: 0.1660
Epoch 2/50
1363/1363 [==============================] - 16s 12ms/step - loss: 1.4245 - categorical_accuracy: 0.5785 - precision_35: 0.1354 - recall_35: 0.9466 - f1_score: 0.1822 - val_loss: 1.5629 - val_categorical_accuracy: 0.5429 - val_precision_35: 0.1310 - val_recall_35: 0.9384 - val_f1_score: 0.1779
Epoch 3/50
1363/1363 [==============================] - 16s 12ms/step - loss: 1.3806 - categorical_accuracy: 0.5969 - precision_35: 0.1349 - recall_35: 0.9495 - f1_score: 0.1939 - val_loss: 1.5681 - val_categorical_accuracy: 0.5305 - val_precision_35: 0.1293 - val_recall_35: 0.9498 - val_f1_score: 0.1903
Epoch 4/50
1363/1363 [==============================] - 16s 12ms/step - loss: 1.3450 - categorical_accuracy: 0.5993 - precision_35: 0.1330 - recall_35: 0.9582 - f1_score: 0.1972 - val_loss: 1.6460 - val_categorical_accuracy: 0.5249 - val_precision_35: 0.1317 - val_recall_35: 0.9453 - val_f1_score: 0.1981
Epoch 5/50
1363/1363 [==============================] - 16s 12ms/step - loss: 1.3236 - categorical_accuracy: 0.6006 - precision_35: 0.1356 - recall_35: 0.9653 - f1_score: 0.2034 - val_loss: 1.6856 - val_categorical_accuracy: 0.4793 - val_precision_35: 0.1304 - val_recall_35: 0.9497 - val_f1_score: 0.1897
Epoch 6/50
1363/1363 [==============================] - 16s 12ms/step - loss: 1.3279 - categorical_accuracy: 0.6006 - precision_35: 0.1320 - recall_35: 0.9591 - f1_score: 0.2073 - val_loss: 1.5592 - val_categorical_accuracy: 0.5241 - val_precision_35: 0.1300 - val_recall_35: 0.9419 - val_f1_score: 0.1897
Epoch 7/50
1363/1363 [==============================] - 16s 12ms/step - loss: 1.2999 - categorical_accuracy: 0.6135 - precision_35: 0.1336 - recall_35: 0.9622 - f1_score: 0.2107 - val_loss: 1.7586 - val_categorical_accuracy: 0.4796 - val_precision_35: 0.1279 - val_recall_35: 0.9395 - val_f1_score: 0.1885

 1 64 30 0.001 189 48 

Epoch 1/50
1367/1367 [==============================] - 17s 12ms/step - loss: 1.7607 - categorical_accuracy: 0.5091 - precision_36: 0.1192 - recall_36: 0.8680 - f1_score: 0.1546 - val_loss: 1.5714 - val_categorical_accuracy: 0.4971 - val_precision_36: 0.1360 - val_recall_36: 0.9386 - val_f1_score: 0.1837
Epoch 2/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.4563 - categorical_accuracy: 0.5720 - precision_36: 0.1307 - recall_36: 0.9420 - f1_score: 0.1804 - val_loss: 1.4834 - val_categorical_accuracy: 0.5649 - val_precision_36: 0.1304 - val_recall_36: 0.9420 - val_f1_score: 0.1913
Epoch 3/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.4000 - categorical_accuracy: 0.5778 - precision_36: 0.1298 - recall_36: 0.9494 - f1_score: 0.1896 - val_loss: 1.4170 - val_categorical_accuracy: 0.6058 - val_precision_36: 0.1297 - val_recall_36: 0.9544 - val_f1_score: 0.1872
Epoch 4/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.3853 - categorical_accuracy: 0.5789 - precision_36: 0.1288 - recall_36: 0.9527 - f1_score: 0.1948 - val_loss: 1.3732 - val_categorical_accuracy: 0.5863 - val_precision_36: 0.1315 - val_recall_36: 0.9626 - val_f1_score: 0.1933
Epoch 5/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.3617 - categorical_accuracy: 0.5855 - precision_36: 0.1293 - recall_36: 0.9547 - f1_score: 0.2025 - val_loss: 1.4197 - val_categorical_accuracy: 0.5869 - val_precision_36: 0.1296 - val_recall_36: 0.9523 - val_f1_score: 0.2000
Epoch 6/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.3575 - categorical_accuracy: 0.5901 - precision_36: 0.1278 - recall_36: 0.9541 - f1_score: 0.2090 - val_loss: 1.3669 - val_categorical_accuracy: 0.5690 - val_precision_36: 0.1281 - val_recall_36: 0.9706 - val_f1_score: 0.1904
Epoch 7/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.3299 - categorical_accuracy: 0.5976 - precision_36: 0.1293 - recall_36: 0.9602 - f1_score: 0.2188 - val_loss: 1.4142 - val_categorical_accuracy: 0.6095 - val_precision_36: 0.1304 - val_recall_36: 0.9632 - val_f1_score: 0.1982
Epoch 8/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.3254 - categorical_accuracy: 0.5957 - precision_36: 0.1302 - recall_36: 0.9596 - f1_score: 0.2168 - val_loss: 1.3974 - val_categorical_accuracy: 0.5936 - val_precision_36: 0.1342 - val_recall_36: 0.9628 - val_f1_score: 0.2034
Epoch 9/50
1367/1367 [==============================] - 16s 12ms/step - loss: 1.2997 - categorical_accuracy: 0.6224 - precision_36: 0.1299 - recall_36: 0.9555 - f1_score: 0.2309 - val_loss: 1.4094 - val_categorical_accuracy: 0.6372 - val_precision_36: 0.1284 - val_recall_36: 0.9320 - val_f1_score: 0.2005
Epoch 10/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.3182 - categorical_accuracy: 0.6039 - precision_36: 0.1312 - recall_36: 0.9555 - f1_score: 0.2295 - val_loss: 1.3724 - val_categorical_accuracy: 0.6209 - val_precision_36: 0.1343 - val_recall_36: 0.9639 - val_f1_score: 0.2066
Epoch 11/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.2883 - categorical_accuracy: 0.6187 - precision_36: 0.1324 - recall_36: 0.9583 - f1_score: 0.2335 - val_loss: 1.4336 - val_categorical_accuracy: 0.5655 - val_precision_36: 0.1323 - val_recall_36: 0.9537 - val_f1_score: 0.2054
Epoch 12/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2962 - categorical_accuracy: 0.6121 - precision_36: 0.1315 - recall_36: 0.9524 - f1_score: 0.2394 - val_loss: 1.3716 - val_categorical_accuracy: 0.6071 - val_precision_36: 0.1327 - val_recall_36: 0.9647 - val_f1_score: 0.2081
Epoch 13/50
1367/1367 [==============================] - 16s 11ms/step - loss: 1.2769 - categorical_accuracy: 0.6113 - precision_36: 0.1326 - recall_36: 0.9533 - f1_score: 0.2380 - val_loss: 1.3453 - val_categorical_accuracy: 0.5889 - val_precision_36: 0.1353 - val_recall_36: 0.9738 - val_f1_score: 0.2184
Epoch 14/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2860 - categorical_accuracy: 0.6126 - precision_36: 0.1315 - recall_36: 0.9519 - f1_score: 0.2405 - val_loss: 1.2934 - val_categorical_accuracy: 0.5847 - val_precision_36: 0.1363 - val_recall_36: 0.9599 - val_f1_score: 0.2141
Epoch 15/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2762 - categorical_accuracy: 0.6150 - precision_36: 0.1332 - recall_36: 0.9497 - f1_score: 0.2402 - val_loss: 1.3019 - val_categorical_accuracy: 0.6152 - val_precision_36: 0.1348 - val_recall_36: 0.9585 - val_f1_score: 0.2032
Epoch 16/50
1367/1367 [==============================] - 15s 11ms/step - loss: 1.2719 - categorical_accuracy: 0.6178 - precision_36: 0.1344 - recall_36: 0.9518 - f1_score: 0.2342 - val_loss: 1.3464 - val_categorical_accuracy: 0.6246 - val_precision_36: 0.1328 - val_recall_36: 0.9444 - val_f1_score: 0.2066

 2 64 30 0.001 190 47 

Epoch 1/50
1302/1302 [==============================] - 17s 13ms/step - loss: 1.6840 - categorical_accuracy: 0.5141 - precision_37: 0.1265 - recall_37: 0.8678 - f1_score: 0.1536 - val_loss: 2.2486 - val_categorical_accuracy: 0.4511 - val_precision_37: 0.1233 - val_recall_37: 0.8543 - val_f1_score: 0.1550
Epoch 2/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.4174 - categorical_accuracy: 0.5732 - precision_37: 0.1360 - recall_37: 0.9393 - f1_score: 0.1833 - val_loss: 2.1374 - val_categorical_accuracy: 0.4410 - val_precision_37: 0.1240 - val_recall_37: 0.8973 - val_f1_score: 0.1542
Epoch 3/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.3642 - categorical_accuracy: 0.5881 - precision_37: 0.1320 - recall_37: 0.9463 - f1_score: 0.1919 - val_loss: 2.1997 - val_categorical_accuracy: 0.4409 - val_precision_37: 0.1186 - val_recall_37: 0.9002 - val_f1_score: 0.1538
Epoch 4/50
1302/1302 [==============================] - 16s 13ms/step - loss: 1.3305 - categorical_accuracy: 0.5940 - precision_37: 0.1331 - recall_37: 0.9507 - f1_score: 0.1918 - val_loss: 2.3175 - val_categorical_accuracy: 0.4254 - val_precision_37: 0.1219 - val_recall_37: 0.8991 - val_f1_score: 0.1598
Epoch 5/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.3201 - categorical_accuracy: 0.5976 - precision_37: 0.1338 - recall_37: 0.9516 - f1_score: 0.2057 - val_loss: 2.3671 - val_categorical_accuracy: 0.4713 - val_precision_37: 0.1243 - val_recall_37: 0.8974 - val_f1_score: 0.1638
Epoch 6/50
1302/1302 [==============================] - 15s 12ms/step - loss: 1.3023 - categorical_accuracy: 0.5954 - precision_37: 0.1349 - recall_37: 0.9540 - f1_score: 0.2085 - val_loss: 2.2928 - val_categorical_accuracy: 0.4594 - val_precision_37: 0.1190 - val_recall_37: 0.8874 - val_f1_score: 0.1567
Epoch 7/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.3077 - categorical_accuracy: 0.5990 - precision_37: 0.1325 - recall_37: 0.9542 - f1_score: 0.2101 - val_loss: 2.3921 - val_categorical_accuracy: 0.4606 - val_precision_37: 0.1213 - val_recall_37: 0.8895 - val_f1_score: 0.1592
Epoch 8/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2983 - categorical_accuracy: 0.6022 - precision_37: 0.1345 - recall_37: 0.9525 - f1_score: 0.2180 - val_loss: 2.2415 - val_categorical_accuracy: 0.4373 - val_precision_37: 0.1251 - val_recall_37: 0.9318 - val_f1_score: 0.1640
Epoch 9/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2868 - categorical_accuracy: 0.6041 - precision_37: 0.1319 - recall_37: 0.9502 - f1_score: 0.2093 - val_loss: 2.2906 - val_categorical_accuracy: 0.4675 - val_precision_37: 0.1203 - val_recall_37: 0.8710 - val_f1_score: 0.1616
Epoch 10/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2607 - categorical_accuracy: 0.6109 - precision_37: 0.1343 - recall_37: 0.9555 - f1_score: 0.2196 - val_loss: 2.2299 - val_categorical_accuracy: 0.4832 - val_precision_37: 0.1256 - val_recall_37: 0.9204 - val_f1_score: 0.1646
Epoch 11/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2702 - categorical_accuracy: 0.6030 - precision_37: 0.1363 - recall_37: 0.9540 - f1_score: 0.2095 - val_loss: 2.3051 - val_categorical_accuracy: 0.4572 - val_precision_37: 0.1194 - val_recall_37: 0.8740 - val_f1_score: 0.1593
Epoch 12/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2556 - categorical_accuracy: 0.6116 - precision_37: 0.1355 - recall_37: 0.9536 - f1_score: 0.2209 - val_loss: 2.3873 - val_categorical_accuracy: 0.4476 - val_precision_37: 0.1255 - val_recall_37: 0.8736 - val_f1_score: 0.1689
Epoch 13/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2603 - categorical_accuracy: 0.6153 - precision_37: 0.1381 - recall_37: 0.9508 - f1_score: 0.2212 - val_loss: 2.2618 - val_categorical_accuracy: 0.4592 - val_precision_37: 0.1251 - val_recall_37: 0.8753 - val_f1_score: 0.1663
Epoch 14/50
1302/1302 [==============================] - 15s 12ms/step - loss: 1.2412 - categorical_accuracy: 0.6148 - precision_37: 0.1380 - recall_37: 0.9515 - f1_score: 0.2280 - val_loss: 2.3480 - val_categorical_accuracy: 0.4617 - val_precision_37: 0.1248 - val_recall_37: 0.8763 - val_f1_score: 0.1666
Epoch 15/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2341 - categorical_accuracy: 0.6231 - precision_37: 0.1384 - recall_37: 0.9487 - f1_score: 0.2308 - val_loss: 2.3455 - val_categorical_accuracy: 0.4867 - val_precision_37: 0.1263 - val_recall_37: 0.8741 - val_f1_score: 0.1712
Epoch 16/50
1302/1302 [==============================] - 15s 12ms/step - loss: 1.2387 - categorical_accuracy: 0.6230 - precision_37: 0.1403 - recall_37: 0.9500 - f1_score: 0.2405 - val_loss: 2.3370 - val_categorical_accuracy: 0.4435 - val_precision_37: 0.1237 - val_recall_37: 0.8531 - val_f1_score: 0.1655
Epoch 17/50
1302/1302 [==============================] - 16s 12ms/step - loss: 1.2310 - categorical_accuracy: 0.6237 - precision_37: 0.1396 - recall_37: 0.9487 - f1_score: 0.2484 - val_loss: 2.3450 - val_categorical_accuracy: 0.4671 - val_precision_37: 0.1306 - val_recall_37: 0.9191 - val_f1_score: 0.1711
Epoch 18/50
1302/1302 [==============================] - 15s 12ms/step - loss: 1.2349 - categorical_accuracy: 0.6249 - precision_37: 0.1395 - recall_37: 0.9479 - f1_score: 0.2416 - val_loss: 2.4740 - val_categorical_accuracy: 0.4554 - val_precision_37: 0.1263 - val_recall_37: 0.8915 - val_f1_score: 0.1666

 3 64 30 0.001 190 47 

Epoch 1/50
1384/1384 [==============================] - 18s 12ms/step - loss: 1.7511 - categorical_accuracy: 0.5004 - precision_38: 0.1221 - recall_38: 0.8683 - f1_score: 0.1601 - val_loss: 1.4534 - val_categorical_accuracy: 0.5907 - val_precision_38: 0.1351 - val_recall_38: 0.9322 - val_f1_score: 0.1781
Epoch 2/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.4545 - categorical_accuracy: 0.5494 - precision_38: 0.1316 - recall_38: 0.9430 - f1_score: 0.1855 - val_loss: 1.4483 - val_categorical_accuracy: 0.5527 - val_precision_38: 0.1308 - val_recall_38: 0.9345 - val_f1_score: 0.1846
Epoch 3/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.4105 - categorical_accuracy: 0.5741 - precision_38: 0.1322 - recall_38: 0.9491 - f1_score: 0.1975 - val_loss: 1.3494 - val_categorical_accuracy: 0.5934 - val_precision_38: 0.1258 - val_recall_38: 0.9387 - val_f1_score: 0.1840
Epoch 4/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.3690 - categorical_accuracy: 0.5803 - precision_38: 0.1324 - recall_38: 0.9569 - f1_score: 0.1986 - val_loss: 1.3934 - val_categorical_accuracy: 0.5660 - val_precision_38: 0.1351 - val_recall_38: 0.9493 - val_f1_score: 0.1933
Epoch 5/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3622 - categorical_accuracy: 0.5855 - precision_38: 0.1309 - recall_38: 0.9615 - f1_score: 0.2028 - val_loss: 1.3424 - val_categorical_accuracy: 0.6108 - val_precision_38: 0.1315 - val_recall_38: 0.9426 - val_f1_score: 0.1994
Epoch 6/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.3332 - categorical_accuracy: 0.5895 - precision_38: 0.1333 - recall_38: 0.9647 - f1_score: 0.2099 - val_loss: 1.4191 - val_categorical_accuracy: 0.6054 - val_precision_38: 0.1297 - val_recall_38: 0.9454 - val_f1_score: 0.2029
Epoch 7/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.3246 - categorical_accuracy: 0.6005 - precision_38: 0.1337 - recall_38: 0.9613 - f1_score: 0.2184 - val_loss: 1.3188 - val_categorical_accuracy: 0.6084 - val_precision_38: 0.1293 - val_recall_38: 0.9472 - val_f1_score: 0.1986
Epoch 8/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.3192 - categorical_accuracy: 0.5958 - precision_38: 0.1330 - recall_38: 0.9644 - f1_score: 0.2171 - val_loss: 1.3251 - val_categorical_accuracy: 0.6025 - val_precision_38: 0.1338 - val_recall_38: 0.9422 - val_f1_score: 0.2010
Epoch 9/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2957 - categorical_accuracy: 0.6015 - precision_38: 0.1346 - recall_38: 0.9629 - f1_score: 0.2279 - val_loss: 1.4146 - val_categorical_accuracy: 0.5337 - val_precision_38: 0.1312 - val_recall_38: 0.9490 - val_f1_score: 0.1912

 4 64 30 0.001 190 47 

Epoch 1/50
1384/1384 [==============================] - 17s 12ms/step - loss: 1.7417 - categorical_accuracy: 0.5049 - precision_39: 0.1263 - recall_39: 0.8767 - f1_score: 0.1644 - val_loss: 1.4306 - val_categorical_accuracy: 0.5654 - val_precision_39: 0.1431 - val_recall_39: 0.9452 - val_f1_score: 0.1772
Epoch 2/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.4604 - categorical_accuracy: 0.5586 - precision_39: 0.1329 - recall_39: 0.9386 - f1_score: 0.1870 - val_loss: 1.4800 - val_categorical_accuracy: 0.5374 - val_precision_39: 0.1331 - val_recall_39: 0.9450 - val_f1_score: 0.1705
Epoch 3/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.4016 - categorical_accuracy: 0.5708 - precision_39: 0.1318 - recall_39: 0.9464 - f1_score: 0.1962 - val_loss: 1.3745 - val_categorical_accuracy: 0.6129 - val_precision_39: 0.1397 - val_recall_39: 0.9601 - val_f1_score: 0.1782
Epoch 4/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3847 - categorical_accuracy: 0.5757 - precision_39: 0.1321 - recall_39: 0.9499 - f1_score: 0.2012 - val_loss: 1.5143 - val_categorical_accuracy: 0.6151 - val_precision_39: 0.1290 - val_recall_39: 0.9388 - val_f1_score: 0.1704
Epoch 5/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.3641 - categorical_accuracy: 0.5911 - precision_39: 0.1319 - recall_39: 0.9479 - f1_score: 0.2117 - val_loss: 1.4278 - val_categorical_accuracy: 0.6083 - val_precision_39: 0.1374 - val_recall_39: 0.9456 - val_f1_score: 0.1860
Epoch 6/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3610 - categorical_accuracy: 0.5881 - precision_39: 0.1313 - recall_39: 0.9526 - f1_score: 0.2112 - val_loss: 1.2696 - val_categorical_accuracy: 0.6520 - val_precision_39: 0.1342 - val_recall_39: 0.9582 - val_f1_score: 0.1866
Epoch 7/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.3436 - categorical_accuracy: 0.5924 - precision_39: 0.1314 - recall_39: 0.9538 - f1_score: 0.2186 - val_loss: 1.3433 - val_categorical_accuracy: 0.6122 - val_precision_39: 0.1328 - val_recall_39: 0.9545 - val_f1_score: 0.1831
Epoch 8/50
1384/1384 [==============================] - 16s 12ms/step - loss: 1.3210 - categorical_accuracy: 0.5971 - precision_39: 0.1325 - recall_39: 0.9563 - f1_score: 0.2279 - val_loss: 1.2966 - val_categorical_accuracy: 0.6456 - val_precision_39: 0.1311 - val_recall_39: 0.9570 - val_f1_score: 0.1798
Epoch 9/50
1384/1384 [==============================] - 17s 12ms/step - loss: 1.3303 - categorical_accuracy: 0.5913 - precision_39: 0.1323 - recall_39: 0.9543 - f1_score: 0.2216 - val_loss: 1.3124 - val_categorical_accuracy: 0.6165 - val_precision_39: 0.1339 - val_recall_39: 0.9586 - val_f1_score: 0.1914
Epoch 10/50
1384/1384 [==============================] - 17s 12ms/step - loss: 1.3142 - categorical_accuracy: 0.5983 - precision_39: 0.1325 - recall_39: 0.9571 - f1_score: 0.2354 - val_loss: 1.3469 - val_categorical_accuracy: 0.5891 - val_precision_39: 0.1324 - val_recall_39: 0.9580 - val_f1_score: 0.1762
Epoch 11/50
1384/1384 [==============================] - 14s 10ms/step - loss: 1.3073 - categorical_accuracy: 0.6093 - precision_39: 0.1336 - recall_39: 0.9517 - f1_score: 0.2460 - val_loss: 1.2693 - val_categorical_accuracy: 0.6227 - val_precision_39: 0.1396 - val_recall_39: 0.9508 - val_f1_score: 0.1924
Epoch 12/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.2949 - categorical_accuracy: 0.6027 - precision_39: 0.1357 - recall_39: 0.9557 - f1_score: 0.2427 - val_loss: 1.2850 - val_categorical_accuracy: 0.6334 - val_precision_39: 0.1347 - val_recall_39: 0.9547 - val_f1_score: 0.1926
Epoch 13/50
1384/1384 [==============================] - 14s 10ms/step - loss: 1.2970 - categorical_accuracy: 0.6000 - precision_39: 0.1342 - recall_39: 0.9537 - f1_score: 0.2409 - val_loss: 1.2687 - val_categorical_accuracy: 0.6406 - val_precision_39: 0.1351 - val_recall_39: 0.9555 - val_f1_score: 0.1933
Epoch 14/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2933 - categorical_accuracy: 0.6116 - precision_39: 0.1358 - recall_39: 0.9495 - f1_score: 0.2483 - val_loss: 1.2182 - val_categorical_accuracy: 0.6512 - val_precision_39: 0.1351 - val_recall_39: 0.9554 - val_f1_score: 0.1859
Epoch 15/50
1384/1384 [==============================] - 14s 10ms/step - loss: 1.2847 - categorical_accuracy: 0.6119 - precision_39: 0.1351 - recall_39: 0.9543 - f1_score: 0.2505 - val_loss: 1.2407 - val_categorical_accuracy: 0.6581 - val_precision_39: 0.1392 - val_recall_39: 0.9555 - val_f1_score: 0.1913
Epoch 16/50
1384/1384 [==============================] - 14s 10ms/step - loss: 1.2734 - categorical_accuracy: 0.6242 - precision_39: 0.1361 - recall_39: 0.9511 - f1_score: 0.2622 - val_loss: 1.3422 - val_categorical_accuracy: 0.5958 - val_precision_39: 0.1379 - val_recall_39: 0.9574 - val_f1_score: 0.2137
Epoch 17/50
1384/1384 [==============================] - 13s 10ms/step - loss: 1.2728 - categorical_accuracy: 0.6139 - precision_39: 0.1375 - recall_39: 0.9495 - f1_score: 0.2626 - val_loss: 1.2775 - val_categorical_accuracy: 0.6400 - val_precision_39: 0.1413 - val_recall_39: 0.9537 - val_f1_score: 0.2038
Epoch 18/50
1384/1384 [==============================] - 16s 11ms/step - loss: 1.2752 - categorical_accuracy: 0.6119 - precision_39: 0.1381 - recall_39: 0.9529 - f1_score: 0.2668 - val_loss: 1.2264 - val_categorical_accuracy: 0.6587 - val_precision_39: 0.1371 - val_recall_39: 0.9485 - val_f1_score: 0.1976
Epoch 19/50
1384/1384 [==============================] - 15s 11ms/step - loss: 1.2693 - categorical_accuracy: 0.6174 - precision_39: 0.1367 - recall_39: 0.9475 - f1_score: 0.2551 - val_loss: 1.2093 - val_categorical_accuracy: 0.6545 - val_precision_39: 0.1406 - val_recall_39: 0.9566 - val_f1_score: 0.2102

			val_accuracy		val_precision		val_recall		val_f1_score
			mean	std	mean	std	mean	std	mean	std
batch_size	seq_length	learning_rate
32	15	0.0001	0.572148	0.073204	0.408438	0.063195	0.396537	0.027559	0.378847	0.040708
		0.0010	0.577516	0.068717	0.476084	0.080048	0.440360	0.029731	0.424277	0.039733
	30	0.0001	0.561612	0.062351	0.474269	0.072688	0.435835	0.039429	0.412130	0.050721
		0.0010	0.551434	0.087456	0.495603	0.113382	0.441186	0.050074	0.419919	0.076330
64	15	0.0001	0.565692	0.058694	0.420889	0.055004	0.407307	0.015851	0.388302	0.020709
		0.0010	0.573203	0.061259	0.452327	0.072715	0.424280	0.022099	0.412525	0.040651
	30	0.0001	0.556785	0.073003	0.461286	0.098666	0.414262	0.054505	0.402358	0.067456
		0.0010	0.560334	0.051917	0.488388	0.100515	0.446361	0.017267	0.426248	0.044673

Wall time: 5h 4min 32s

rep_df = pd.DataFrame(linear_param_fit.report)
rep_df['val_accuracy'] = rep_df.class_rep_val.apply(pd.Series)['accuracy']
rep_df['val_precision'] = rep_df.class_rep_val.apply(pd.Series)['macro avg'].apply(pd.Series)['precision']
rep_df['val_recall'] = rep_df.class_rep_val.apply(pd.Series)['macro avg'].apply(pd.Series)['recall']
rep_df['val_f1_score'] = rep_df.class_rep_val.apply(pd.Series)['macro avg'].apply(pd.Series)['f1-score']
rep_df.drop(['class_rep_val', 'class_rep_test'], axis=1, inplace=True)

fdict = {'val_accuracy': ['mean', 'std'], 'val_precision': ['mean', 'std'],
         'val_recall': ['mean', 'std'], 'val_f1_score': ['mean', 'std']}
rep_df_stat = rep_df.groupby(['batch_size', 'seq_length', 'learning_rate']).agg(fdict)      
rep_df_stat.reset_index(inplace=True)

fig, ax = plt.subplots()

plt.style.use('ggplot')
sns.scatterplot(data=rep_df, x='learning_rate', y='val_f1_score',
                    hue=rep_df['batch_size'], palette=['red', 'blue'], alpha=0.25,
                    style=rep_df['seq_length'], ax=ax)

sns.scatterplot(data=rep_df_stat, x='learning_rate', y=('val_f1_score', 'mean'),
                    hue='batch_size', palette=['red', 'blue'],
                    style='seq_length', ax=ax, s=100)

#rep_df_stat.plot.scatter(x='learning_rate', y=('val_f1_score', 'mean'), ax=ax, s=100, marker='x')
h, l = ax.get_legend_handles_labels()
ax.legend(handles=h[:-6], labels=l[:-6], loc='lower center')

plt.show()

%%time

train_df, test_df, val_df = dset.split(real=False, simul=True, drawn=True, test_size=0.2, val_size=0.1)

tf.keras.backend.clear_session()
tf.compat.v1.reset_default_graph()

linear_model = kmodel('LinearModel', LinearModel, flist0, 'class', categories, 64, 30, 
                      train_df, val_df, test_df, 'D:/datatmp', reset_ts=False, class_bal=True)

linear_model.compile_and_fit(max_epochs=50, patience=5, lr=0.001)

linear_model.save()

Instances Train: 690  Test: 192  Validation: 77
Model: "linear_model_40"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 flatten_40 (Flatten)        (64, 480)                 0         
                                                                 
 layer_normalization_40 (Lay  (64, 480)                960       
 erNormalization)                                                
                                                                 
 dense_40 (Dense)            (64, 17)                  8177      
                                                                 
=================================================================
Total params: 9,137
Trainable params: 9,137
Non-trainable params: 0
_________________________________________________________________
None
Epoch 1/50
7095/7095 [==============================] - 159s 22ms/step - loss: 1.2580 - categorical_accuracy: 0.6088 - precision_40: 0.1208 - recall_40: 0.9416 - f1_score: 0.1702 - val_loss: 1.1321 - val_categorical_accuracy: 0.6807 - val_precision_40: 0.1263 - val_recall_40: 0.9643 - val_f1_score: 0.1799
Epoch 2/50
7095/7095 [==============================] - 80s 11ms/step - loss: 0.9494 - categorical_accuracy: 0.6529 - precision_40: 0.1242 - recall_40: 0.9767 - f1_score: 0.1753 - val_loss: 0.9508 - val_categorical_accuracy: 0.7281 - val_precision_40: 0.1235 - val_recall_40: 0.9849 - val_f1_score: 0.1768
Epoch 3/50
7095/7095 [==============================] - 79s 11ms/step - loss: 0.9184 - categorical_accuracy: 0.6583 - precision_40: 0.1255 - recall_40: 0.9786 - f1_score: 0.1774 - val_loss: 0.8233 - val_categorical_accuracy: 0.7722 - val_precision_40: 0.1268 - val_recall_40: 0.9830 - val_f1_score: 0.1844
Epoch 4/50
7095/7095 [==============================] - 82s 12ms/step - loss: 0.8818 - categorical_accuracy: 0.6649 - precision_40: 0.1265 - recall_40: 0.9776 - f1_score: 0.1795 - val_loss: 0.8927 - val_categorical_accuracy: 0.7243 - val_precision_40: 0.1287 - val_recall_40: 0.9811 - val_f1_score: 0.1733
Epoch 5/50
7095/7095 [==============================] - 80s 11ms/step - loss: 0.8411 - categorical_accuracy: 0.6761 - precision_40: 0.1287 - recall_40: 0.9765 - f1_score: 0.1769 - val_loss: 0.8744 - val_categorical_accuracy: 0.7430 - val_precision_40: 0.1295 - val_recall_40: 0.9860 - val_f1_score: 0.1875
Epoch 6/50
7095/7095 [==============================] - 80s 11ms/step - loss: 0.8184 - categorical_accuracy: 0.6828 - precision_40: 0.1318 - recall_40: 0.9759 - f1_score: 0.1845 - val_loss: 0.9267 - val_categorical_accuracy: 0.6771 - val_precision_40: 0.1365 - val_recall_40: 0.9839 - val_f1_score: 0.1924
Epoch 7/50
7095/7095 [==============================] - 79s 11ms/step - loss: 0.8197 - categorical_accuracy: 0.6754 - precision_40: 0.1355 - recall_40: 0.9681 - f1_score: 0.1887 - val_loss: 0.8432 - val_categorical_accuracy: 0.7406 - val_precision_40: 0.1380 - val_recall_40: 0.9696 - val_f1_score: 0.1965
Epoch 8/50
7095/7095 [==============================] - 80s 11ms/step - loss: 0.8027 - categorical_accuracy: 0.6811 - precision_40: 0.1382 - recall_40: 0.9681 - f1_score: 0.1945 - val_loss: 0.8929 - val_categorical_accuracy: 0.7484 - val_precision_40: 0.1397 - val_recall_40: 0.9652 - val_f1_score: 0.1976
Epoch 9/50
7095/7095 [==============================] - 79s 11ms/step - loss: 0.7931 - categorical_accuracy: 0.6888 - precision_40: 0.1397 - recall_40: 0.9631 - f1_score: 0.1949 - val_loss: 0.9717 - val_categorical_accuracy: 0.6976 - val_precision_40: 0.1394 - val_recall_40: 0.9748 - val_f1_score: 0.1950
Epoch 10/50
7095/7095 [==============================] - 80s 11ms/step - loss: 0.7881 - categorical_accuracy: 0.6855 - precision_40: 0.1424 - recall_40: 0.9621 - f1_score: 0.1983 - val_loss: 0.8069 - val_categorical_accuracy: 0.7675 - val_precision_40: 0.1461 - val_recall_40: 0.9778 - val_f1_score: 0.2046
Epoch 11/50
7095/7095 [==============================] - 79s 11ms/step - loss: 0.7803 - categorical_accuracy: 0.6866 - precision_40: 0.1451 - recall_40: 0.9570 - f1_score: 0.2029 - val_loss: 0.8575 - val_categorical_accuracy: 0.7324 - val_precision_40: 0.1466 - val_recall_40: 0.9665 - val_f1_score: 0.2046
Epoch 12/50
7095/7095 [==============================] - 82s 11ms/step - loss: 0.7732 - categorical_accuracy: 0.6912 - precision_40: 0.1473 - recall_40: 0.9504 - f1_score: 0.2060 - val_loss: 0.7982 - val_categorical_accuracy: 0.7495 - val_precision_40: 0.1516 - val_recall_40: 0.9666 - val_f1_score: 0.2174
Epoch 13/50
7095/7095 [==============================] - 82s 12ms/step - loss: 0.7773 - categorical_accuracy: 0.6887 - precision_40: 0.1490 - recall_40: 0.9450 - f1_score: 0.2072 - val_loss: 0.9464 - val_categorical_accuracy: 0.7171 - val_precision_40: 0.1469 - val_recall_40: 0.9648 - val_f1_score: 0.2060
Epoch 14/50
7095/7095 [==============================] - 83s 12ms/step - loss: 0.7667 - categorical_accuracy: 0.6884 - precision_40: 0.1504 - recall_40: 0.9412 - f1_score: 0.2077 - val_loss: 0.8710 - val_categorical_accuracy: 0.7220 - val_precision_40: 0.1499 - val_recall_40: 0.9486 - val_f1_score: 0.2050
Epoch 15/50
7095/7095 [==============================] - 78s 11ms/step - loss: 0.7675 - categorical_accuracy: 0.6889 - precision_40: 0.1536 - recall_40: 0.9304 - f1_score: 0.2126 - val_loss: 0.9404 - val_categorical_accuracy: 0.7119 - val_precision_40: 0.1520 - val_recall_40: 0.9390 - val_f1_score: 0.2103
Epoch 16/50
7095/7095 [==============================] - 77s 11ms/step - loss: 0.7677 - categorical_accuracy: 0.6911 - precision_40: 0.1544 - recall_40: 0.9210 - f1_score: 0.2119 - val_loss: 0.9921 - val_categorical_accuracy: 0.7138 - val_precision_40: 0.1545 - val_recall_40: 0.9446 - val_f1_score: 0.2095
Epoch 17/50
7095/7095 [==============================] - 79s 11ms/step - loss: 0.7605 - categorical_accuracy: 0.6903 - precision_40: 0.1558 - recall_40: 0.9136 - f1_score: 0.2138 - val_loss: 0.8364 - val_categorical_accuracy: 0.7325 - val_precision_40: 0.1573 - val_recall_40: 0.9410 - val_f1_score: 0.2150

Val Classification Report
              precision    recall  f1-score   support

           0       0.26      0.12      0.16       880
           1       0.81      0.82      0.81      4200
           2       0.95      1.00      0.97       360
           3       0.80      0.14      0.24      6030
           5       0.99      0.90      0.94     14062
           6       0.98      0.99      0.99      5100
           7       0.29      0.59      0.39        87
           8       0.81      0.95      0.88      1148
         101       0.76      0.74      0.75      6000
         102       0.57      0.98      0.72        60
         105       0.75      0.95      0.84      2038
         106       0.73      0.91      0.81      2040
         107       0.17      0.68      0.28      1210
         108       0.29      0.59      0.39      1792

    accuracy                           0.75     45007
   macro avg       0.65      0.74      0.65     45007
weighted avg       0.82      0.75      0.75     45007

Test Classification Report
              precision    recall  f1-score   support

           0       0.24      0.46      0.32      2424
           1       0.68      0.84      0.75      9129
           2       0.61      0.96      0.74      1080
           3       0.79      0.23      0.35     13674
           5       0.92      0.83      0.87     35281
           6       0.99      0.97      0.98     12867
           7       0.02      0.01      0.02       217
           8       0.82      0.95      0.88      2163
         101       0.70      0.40      0.51     19782
         102       0.26      0.96      0.41       180
         105       0.78      0.95      0.85      5199
         106       0.67      0.93      0.78      5193
         107       0.13      0.47      0.20      3099
         108       0.29      0.60      0.39      4557

    accuracy                           0.69    114845
   macro avg       0.56      0.68      0.58    114845
weighted avg       0.77      0.69      0.69    114845

All Data Classification Report
              precision    recall  f1-score   support

           0       0.40      0.46      0.43     14210
           1       0.72      0.88      0.79     48336
           2       0.75      0.98      0.85      5760
           3       0.76      0.19      0.30     68980
           5       0.96      0.85      0.90    176014
           6       0.97      0.97      0.97     64434
           7       0.66      0.72      0.69      1801
           8       0.76      0.96      0.85     10059
         101       0.69      0.53      0.60     86657
         102       0.35      0.97      0.52       960
         105       0.83      0.95      0.88     25926
         106       0.69      0.92      0.79     25866
         107       0.31      0.66      0.42     34072
         108       0.31      0.61      0.41     23961

    accuracy                           0.72    587036
   macro avg       0.66      0.76      0.67    587036
weighted avg       0.78      0.72      0.72    587036

model weights saved in linear_model_40_batch_size_64-seq_length_30-2022-09-08T16_28_24.h5
Wall time: 26min 30s

DNN model

class DNNModel(tf.keras.Model):

  def __init__(self):
    super().__init__()
    self.flatten = layers.Flatten()
    self.norm = layers.LayerNormalization()
    self.dense0 = layers.Dense(units=256, activation='relu')
    self.drop0 = layers.Dropout(0.1, seed=200560)
    self.dense1 = layers.Dense(units=128, activation='relu')
    self.drop1 = layers.Dropout(0.1, seed=200560)
    self.dense2 = layers.Dense(units=64, activation='relu')
    self.drop2 = layers.Dropout(0.1, seed=200560)
    self.dense3 = layers.Dense(units=17, activation='relu')
    self.softmax = layers.Softmax()

  def call(self, inputs, training=False):
    x = self.flatten(inputs)
    x = self.norm(x)
    x = self.dense0(x)
    if training:
        x = self.drop0(x, training=training)
    x = self.dense1(x)
    if training:
        x = self.drop1(x, training=training)
    x = self.dense2(x)
    if training:
        x = self.drop2(x, training=training)
    x = self.dense3(x)
    return self.softmax(x)

%%time

train_df, test_df, val_df = dset.split(real=True, simul=True, drawn=True, test_size=0.2, val_size=0.1)

tf.keras.backend.clear_session()
tf.compat.v1.reset_default_graph()

dnn_model = kmodel('DNNModel', DNNModel, flist0, 'class', categories, 64, 30, 
                      train_df, val_df, test_df, 'D:/datatmp', reset_ts=False, class_bal=True)

dnn_model.compile_and_fit(max_epochs=50, patience=5, lr=0.0001)

dnn_model.save()

Instances Train: 1425  Test: 397  Validation: 159
Model: "dnn_model"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 flatten (Flatten)           (64, 480)                 0         
                                                                 
 layer_normalization (LayerN  (64, 480)                960       
 ormalization)                                                   
                                                                 
 dense (Dense)               (64, 256)                 123136    
                                                                 
 dropout (Dropout)           (64, 256)                 0         
                                                                 
 dense_1 (Dense)             (64, 128)                 32896     
                                                                 
 dropout_1 (Dropout)         (64, 128)                 0         
                                                                 
 dense_2 (Dense)             (64, 64)                  8256      
                                                                 
 dropout_2 (Dropout)         (64, 64)                  0         
                                                                 
 dense_3 (Dense)             (64, 17)                  1105      
                                                                 
 softmax (Softmax)           (64, 17)                  0         
                                                                 
=================================================================
Total params: 166,353
Trainable params: 166,353
Non-trainable params: 0
_________________________________________________________________
None
Epoch 1/50
9819/9819 [==============================] - 122s 12ms/step - loss: 1.3475 - categorical_accuracy: 0.5779 - precision: 0.8182 - recall: 0.4472 - f1_score: 0.4466 - val_loss: 0.9733 - val_categorical_accuracy: 0.6828 - val_precision: 0.8273 - val_recall: 0.5979 - val_f1_score: 0.5242
Epoch 2/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.7546 - categorical_accuracy: 0.7031 - precision: 0.8311 - recall: 0.6391 - f1_score: 0.5945 - val_loss: 0.7486 - val_categorical_accuracy: 0.7485 - val_precision: 0.8556 - val_recall: 0.7000 - val_f1_score: 0.5966
Epoch 3/50
9819/9819 [==============================] - 95s 10ms/step - loss: 0.5795 - categorical_accuracy: 0.7492 - precision: 0.8525 - recall: 0.7041 - f1_score: 0.6410 - val_loss: 0.7118 - val_categorical_accuracy: 0.7595 - val_precision: 0.8687 - val_recall: 0.7113 - val_f1_score: 0.6026
Epoch 4/50
9819/9819 [==============================] - 100s 10ms/step - loss: 0.5139 - categorical_accuracy: 0.7724 - precision: 0.8683 - recall: 0.7363 - f1_score: 0.6603 - val_loss: 0.6200 - val_categorical_accuracy: 0.7947 - val_precision: 0.8909 - val_recall: 0.7662 - val_f1_score: 0.6279
Epoch 5/50
9819/9819 [==============================] - 96s 10ms/step - loss: 0.4886 - categorical_accuracy: 0.7886 - precision: 0.8794 - recall: 0.7589 - f1_score: 0.6746 - val_loss: 0.6038 - val_categorical_accuracy: 0.8016 - val_precision: 0.8938 - val_recall: 0.7745 - val_f1_score: 0.6279
Epoch 6/50
9819/9819 [==============================] - 99s 10ms/step - loss: 0.4588 - categorical_accuracy: 0.7983 - precision: 0.8859 - recall: 0.7729 - f1_score: 0.6825 - val_loss: 0.5691 - val_categorical_accuracy: 0.8053 - val_precision: 0.8988 - val_recall: 0.7817 - val_f1_score: 0.6301
Epoch 7/50
9819/9819 [==============================] - 101s 10ms/step - loss: 0.4438 - categorical_accuracy: 0.8082 - precision: 0.8935 - recall: 0.7872 - f1_score: 0.6910 - val_loss: 0.5825 - val_categorical_accuracy: 0.7996 - val_precision: 0.8939 - val_recall: 0.7760 - val_f1_score: 0.6454
Epoch 8/50
9819/9819 [==============================] - 99s 10ms/step - loss: 0.4214 - categorical_accuracy: 0.8154 - precision: 0.8996 - recall: 0.7970 - f1_score: 0.6989 - val_loss: 0.5398 - val_categorical_accuracy: 0.8133 - val_precision: 0.9045 - val_recall: 0.7989 - val_f1_score: 0.6464
Epoch 9/50
9819/9819 [==============================] - 100s 10ms/step - loss: 0.4015 - categorical_accuracy: 0.8207 - precision: 0.9033 - recall: 0.8050 - f1_score: 0.7059 - val_loss: 0.5195 - val_categorical_accuracy: 0.8216 - val_precision: 0.9138 - val_recall: 0.8104 - val_f1_score: 0.6527
Epoch 10/50
9819/9819 [==============================] - 137s 14ms/step - loss: 0.3963 - categorical_accuracy: 0.8283 - precision: 0.9103 - recall: 0.8139 - f1_score: 0.7132 - val_loss: 0.5231 - val_categorical_accuracy: 0.8204 - val_precision: 0.9088 - val_recall: 0.8118 - val_f1_score: 0.6540
Epoch 11/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.3853 - categorical_accuracy: 0.8314 - precision: 0.9125 - recall: 0.8190 - f1_score: 0.7153 - val_loss: 0.5043 - val_categorical_accuracy: 0.8237 - val_precision: 0.9142 - val_recall: 0.8128 - val_f1_score: 0.6636
Epoch 12/50
9819/9819 [==============================] - 98s 10ms/step - loss: 0.3792 - categorical_accuracy: 0.8361 - precision: 0.9168 - recall: 0.8248 - f1_score: 0.7220 - val_loss: 0.4975 - val_categorical_accuracy: 0.8294 - val_precision: 0.9212 - val_recall: 0.8168 - val_f1_score: 0.6641
Epoch 13/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3672 - categorical_accuracy: 0.8397 - precision: 0.9200 - recall: 0.8296 - f1_score: 0.7241 - val_loss: 0.5210 - val_categorical_accuracy: 0.8217 - val_precision: 0.9091 - val_recall: 0.8131 - val_f1_score: 0.6603
Epoch 14/50
9819/9819 [==============================] - 94s 10ms/step - loss: 0.3544 - categorical_accuracy: 0.8441 - precision: 0.9240 - recall: 0.8352 - f1_score: 0.7322 - val_loss: 0.4848 - val_categorical_accuracy: 0.8347 - val_precision: 0.9224 - val_recall: 0.8284 - val_f1_score: 0.6614
Epoch 15/50
9819/9819 [==============================] - 99s 10ms/step - loss: 0.3541 - categorical_accuracy: 0.8457 - precision: 0.9254 - recall: 0.8377 - f1_score: 0.7313 - val_loss: 0.4851 - val_categorical_accuracy: 0.8329 - val_precision: 0.9232 - val_recall: 0.8231 - val_f1_score: 0.6686
Epoch 16/50
9819/9819 [==============================] - 94s 10ms/step - loss: 0.3447 - categorical_accuracy: 0.8489 - precision: 0.9280 - recall: 0.8409 - f1_score: 0.7344 - val_loss: 0.4555 - val_categorical_accuracy: 0.8400 - val_precision: 0.9279 - val_recall: 0.8348 - val_f1_score: 0.6799
Epoch 17/50
9819/9819 [==============================] - 98s 10ms/step - loss: 0.3445 - categorical_accuracy: 0.8503 - precision: 0.9291 - recall: 0.8432 - f1_score: 0.7347 - val_loss: 0.4761 - val_categorical_accuracy: 0.8352 - val_precision: 0.9234 - val_recall: 0.8299 - val_f1_score: 0.6731
Epoch 18/50
9819/9819 [==============================] - 98s 10ms/step - loss: 0.3422 - categorical_accuracy: 0.8525 - precision: 0.9311 - recall: 0.8458 - f1_score: 0.7381 - val_loss: 0.4777 - val_categorical_accuracy: 0.8359 - val_precision: 0.9223 - val_recall: 0.8320 - val_f1_score: 0.6802
Epoch 19/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3349 - categorical_accuracy: 0.8543 - precision: 0.9328 - recall: 0.8483 - f1_score: 0.7421 - val_loss: 0.4727 - val_categorical_accuracy: 0.8339 - val_precision: 0.9223 - val_recall: 0.8271 - val_f1_score: 0.6650
Epoch 20/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3309 - categorical_accuracy: 0.8562 - precision: 0.9347 - recall: 0.8503 - f1_score: 0.7389 - val_loss: 0.4718 - val_categorical_accuracy: 0.8374 - val_precision: 0.9255 - val_recall: 0.8329 - val_f1_score: 0.6821
Epoch 21/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3273 - categorical_accuracy: 0.8576 - precision: 0.9361 - recall: 0.8518 - f1_score: 0.7439 - val_loss: 0.4648 - val_categorical_accuracy: 0.8395 - val_precision: 0.9282 - val_recall: 0.8340 - val_f1_score: 0.6730
Epoch 22/50
9819/9819 [==============================] - 98s 10ms/step - loss: 0.3193 - categorical_accuracy: 0.8599 - precision: 0.9383 - recall: 0.8550 - f1_score: 0.7467 - val_loss: 0.4601 - val_categorical_accuracy: 0.8397 - val_precision: 0.9255 - val_recall: 0.8369 - val_f1_score: 0.6772
Epoch 23/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3163 - categorical_accuracy: 0.8614 - precision: 0.9396 - recall: 0.8568 - f1_score: 0.7491 - val_loss: 0.4521 - val_categorical_accuracy: 0.8444 - val_precision: 0.9308 - val_recall: 0.8421 - val_f1_score: 0.6812
Epoch 24/50
9819/9819 [==============================] - 96s 10ms/step - loss: 0.3142 - categorical_accuracy: 0.8626 - precision: 0.9407 - recall: 0.8582 - f1_score: 0.7510 - val_loss: 0.4462 - val_categorical_accuracy: 0.8458 - val_precision: 0.9333 - val_recall: 0.8426 - val_f1_score: 0.6821
Epoch 25/50
9819/9819 [==============================] - 100s 10ms/step - loss: 0.3125 - categorical_accuracy: 0.8634 - precision: 0.9413 - recall: 0.8591 - f1_score: 0.7525 - val_loss: 0.4521 - val_categorical_accuracy: 0.8428 - val_precision: 0.9299 - val_recall: 0.8393 - val_f1_score: 0.6823
Epoch 26/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3130 - categorical_accuracy: 0.8648 - precision: 0.9428 - recall: 0.8605 - f1_score: 0.7507 - val_loss: 0.4513 - val_categorical_accuracy: 0.8447 - val_precision: 0.9303 - val_recall: 0.8420 - val_f1_score: 0.6844
Epoch 27/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3050 - categorical_accuracy: 0.8662 - precision: 0.9441 - recall: 0.8625 - f1_score: 0.7574 - val_loss: 0.4413 - val_categorical_accuracy: 0.8477 - val_precision: 0.9341 - val_recall: 0.8456 - val_f1_score: 0.6935
Epoch 28/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3060 - categorical_accuracy: 0.8667 - precision: 0.9446 - recall: 0.8631 - f1_score: 0.7570 - val_loss: 0.4538 - val_categorical_accuracy: 0.8407 - val_precision: 0.9280 - val_recall: 0.8368 - val_f1_score: 0.6817
Epoch 29/50
9819/9819 [==============================] - 98s 10ms/step - loss: 0.2991 - categorical_accuracy: 0.8683 - precision: 0.9463 - recall: 0.8644 - f1_score: 0.7593 - val_loss: 0.4472 - val_categorical_accuracy: 0.8456 - val_precision: 0.9316 - val_recall: 0.8436 - val_f1_score: 0.6766
Epoch 30/50
9819/9819 [==============================] - 98s 10ms/step - loss: 0.3004 - categorical_accuracy: 0.8691 - precision: 0.9471 - recall: 0.8658 - f1_score: 0.7585 - val_loss: 0.4547 - val_categorical_accuracy: 0.8486 - val_precision: 0.9349 - val_recall: 0.8467 - val_f1_score: 0.6939
Epoch 31/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.3024 - categorical_accuracy: 0.8702 - precision: 0.9480 - recall: 0.8671 - f1_score: 0.7590 - val_loss: 0.4303 - val_categorical_accuracy: 0.8501 - val_precision: 0.9361 - val_recall: 0.8481 - val_f1_score: 0.6865
Epoch 32/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.2951 - categorical_accuracy: 0.8709 - precision: 0.9488 - recall: 0.8675 - f1_score: 0.7616 - val_loss: 0.4346 - val_categorical_accuracy: 0.8467 - val_precision: 0.9334 - val_recall: 0.8443 - val_f1_score: 0.6794
Epoch 33/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.2936 - categorical_accuracy: 0.8717 - precision: 0.9496 - recall: 0.8685 - f1_score: 0.7634 - val_loss: 0.4584 - val_categorical_accuracy: 0.8431 - val_precision: 0.9287 - val_recall: 0.8412 - val_f1_score: 0.6878
Epoch 34/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.2900 - categorical_accuracy: 0.8731 - precision: 0.9508 - recall: 0.8702 - f1_score: 0.7622 - val_loss: 0.4592 - val_categorical_accuracy: 0.8397 - val_precision: 0.9262 - val_recall: 0.8375 - val_f1_score: 0.6827
Epoch 35/50
9819/9819 [==============================] - 97s 10ms/step - loss: 0.2871 - categorical_accuracy: 0.8730 - precision: 0.9509 - recall: 0.8702 - f1_score: 0.7631 - val_loss: 0.4421 - val_categorical_accuracy: 0.8541 - val_precision: 0.9416 - val_recall: 0.8522 - val_f1_score: 0.6888

Val Classification Report
              precision    recall  f1-score   support

           0       0.68      0.91      0.78     12978
           1       0.78      0.99      0.87      3780
           2       1.00      0.97      0.98       360
           3       0.99      0.95      0.97      4650
           4       0.80      0.91      0.85      2539
           5       0.99      0.98      0.99     14123
           6       0.00      0.00      0.00      5400
           7       0.00      0.00      0.00        14
           8       0.90      0.99      0.95       748
         101       0.95      0.83      0.89      7435
         102       0.97      0.81      0.89        81
         105       0.88      0.99      0.93      2004
         106       0.84      1.00      0.91      2160
         107       0.55      0.72      0.62       635
         108       0.99      0.95      0.97      1772

    accuracy                           0.85     58679
   macro avg       0.75      0.80      0.77     58679
weighted avg       0.79      0.85      0.81     58679

Test Classification Report
              precision    recall  f1-score   support

           0       0.69      0.88      0.77     33789
           1       0.83      0.98      0.90      9399
           2       0.90      1.00      0.95      1097
           3       0.99      0.93      0.96     15400
           4       0.75      0.97      0.84      6246
           5       1.00      0.97      0.98     35960
           6       0.00      0.00      0.00     12914
           7       0.51      1.00      0.68        87
           8       0.81      0.99      0.89      1746
         101       0.92      0.86      0.89     18727
         102       0.71      0.72      0.72       452
         105       0.82      0.98      0.89      5151
         106       0.83      0.99      0.90      5173
         107       0.91      0.99      0.95     11580
         108       0.99      0.92      0.95      4974

    accuracy                           0.86    162695
   macro avg       0.78      0.88      0.82    162695
weighted avg       0.80      0.86      0.83    162695

All Data Classification Report
              precision    recall  f1-score   support

           0       0.70      0.92      0.80    164820
           1       0.87      0.98      0.92     48436
           2       0.93      1.00      0.96      6042
           3       0.99      0.98      0.99     77554
           4       0.84      0.98      0.91     31211
           5       1.00      0.98      0.99    176186
           6       0.00      0.00      0.00     64648
           7       0.66      0.99      0.79      1815
           8       0.90      0.99      0.94     10059
         101       0.95      0.89      0.92     87937
         102       0.78      0.82      0.80      2339
         105       0.89      0.99      0.94     31150
         106       0.84      1.00      0.91     25955
         107       0.91      0.96      0.93     38338
         108       0.99      0.95      0.97     23961

    accuracy                           0.88    790451
   macro avg       0.82      0.90      0.85    790451
weighted avg       0.82      0.88      0.84    790451

model weights saved in dnn_model_batch_size_64-seq_length_30-2022-09-08T17_33_24.h5
Wall time: 1h 2min 36s

RNN model

class RNNModel(tf.keras.Model):

  def __init__(self):
    super().__init__()
    self.lstm = layers.LSTM(128, return_sequences=False, dropout=0.15, name='lstm')
    self.dense1 = layers.Dense(17, activation='relu', name='dense1')
    self.softmax = layers.Softmax(name='softmax')
    return

  def call(self, inputs, training=False):
    x = self.lstm(inputs, training=training)
    x = self.dense1(x)
    return self.softmax(x)

%%time

train_df, test_df, val_df = dset.split(real=True, simul=True, drawn=True, test_size=0.2, val_size=0.1)

tf.keras.backend.clear_session()
tf.compat.v1.reset_default_graph()

rnn_model = kmodel('RNNModel', RNNModel, flist0, 'class', categories, 64, 30, 
                      train_df, val_df, test_df, 'D:/datatmp', reset_ts=False, class_bal=True)

rnn_model.compile_and_fit(max_epochs=50, patience=5, lr=0.001)

rnn_model.save()

Instances Train: 1425  Test: 397  Validation: 159
Model: "rnn_model"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 lstm (LSTM)                 (64, 128)                 74240     
                                                                 
 dense1 (Dense)              (64, 17)                  2193      
                                                                 
 softmax (Softmax)           (64, 17)                  0         
                                                                 
=================================================================
Total params: 76,433
Trainable params: 76,433
Non-trainable params: 0
_________________________________________________________________
None
Epoch 1/50
9819/9819 [==============================] - 114s 11ms/step - loss: 1.3668 - categorical_accuracy: 0.6274 - precision: 0.7926 - recall: 0.5006 - f1_score: 0.4497 - val_loss: 0.5997 - val_categorical_accuracy: 0.7978 - val_precision: 0.8589 - val_recall: 0.7180 - val_f1_score: 0.5834
Epoch 2/50
9819/9819 [==============================] - 99s 10ms/step - loss: 0.6433 - categorical_accuracy: 0.8318 - precision: 0.8749 - recall: 0.7839 - f1_score: 0.6502 - val_loss: 0.4087 - val_categorical_accuracy: 0.8727 - val_precision: 0.9008 - val_recall: 0.8453 - val_f1_score: 0.6625
Epoch 3/50
9819/9819 [==============================] - 100s 10ms/step - loss: 0.4909 - categorical_accuracy: 0.8868 - precision: 0.9102 - recall: 0.8634 - f1_score: 0.7056 - val_loss: 0.3839 - val_categorical_accuracy: 0.8710 - val_precision: 0.8949 - val_recall: 0.8451 - val_f1_score: 0.6703
Epoch 4/50
9819/9819 [==============================] - 105s 11ms/step - loss: 0.4168 - categorical_accuracy: 0.9084 - precision: 0.9249 - recall: 0.8949 - f1_score: 0.7287 - val_loss: 0.3131 - val_categorical_accuracy: 0.8936 - val_precision: 0.9084 - val_recall: 0.8779 - val_f1_score: 0.7042
Epoch 5/50
9819/9819 [==============================] - 106s 11ms/step - loss: 0.3817 - categorical_accuracy: 0.9222 - precision: 0.9353 - recall: 0.9122 - f1_score: 0.7437 - val_loss: 0.2156 - val_categorical_accuracy: 0.9257 - val_precision: 0.9342 - val_recall: 0.9202 - val_f1_score: 0.7318
Epoch 6/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.3526 - categorical_accuracy: 0.9324 - precision: 0.9424 - recall: 0.9243 - f1_score: 0.7536 - val_loss: 0.2890 - val_categorical_accuracy: 0.9034 - val_precision: 0.9195 - val_recall: 0.8912 - val_f1_score: 0.6519
Epoch 7/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.3336 - categorical_accuracy: 0.9397 - precision: 0.9481 - recall: 0.9327 - f1_score: 0.7598 - val_loss: 0.2302 - val_categorical_accuracy: 0.9238 - val_precision: 0.9285 - val_recall: 0.9205 - val_f1_score: 0.7337
Epoch 8/50
9819/9819 [==============================] - 103s 10ms/step - loss: 0.3138 - categorical_accuracy: 0.9455 - precision: 0.9521 - recall: 0.9405 - f1_score: 0.7681 - val_loss: 0.2043 - val_categorical_accuracy: 0.9295 - val_precision: 0.9348 - val_recall: 0.9261 - val_f1_score: 0.7369
Epoch 9/50
9819/9819 [==============================] - 103s 10ms/step - loss: 0.3114 - categorical_accuracy: 0.9479 - precision: 0.9550 - recall: 0.9435 - f1_score: 0.7701 - val_loss: 0.2375 - val_categorical_accuracy: 0.9289 - val_precision: 0.9341 - val_recall: 0.9224 - val_f1_score: 0.7263
Epoch 10/50
9819/9819 [==============================] - 103s 11ms/step - loss: 0.2971 - categorical_accuracy: 0.9545 - precision: 0.9599 - recall: 0.9511 - f1_score: 0.7736 - val_loss: 0.2579 - val_categorical_accuracy: 0.9187 - val_precision: 0.9252 - val_recall: 0.9139 - val_f1_score: 0.7094
Epoch 11/50
9819/9819 [==============================] - 106s 11ms/step - loss: 0.2903 - categorical_accuracy: 0.9546 - precision: 0.9597 - recall: 0.9521 - f1_score: 0.7765 - val_loss: 0.2144 - val_categorical_accuracy: 0.9266 - val_precision: 0.9335 - val_recall: 0.9217 - val_f1_score: 0.7292
Epoch 12/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.2793 - categorical_accuracy: 0.9589 - precision: 0.9638 - recall: 0.9565 - f1_score: 0.7815 - val_loss: 0.2043 - val_categorical_accuracy: 0.9341 - val_precision: 0.9384 - val_recall: 0.9308 - val_f1_score: 0.7482
Epoch 13/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2800 - categorical_accuracy: 0.9611 - precision: 0.9653 - recall: 0.9588 - f1_score: 0.7819 - val_loss: 0.2424 - val_categorical_accuracy: 0.9295 - val_precision: 0.9332 - val_recall: 0.9263 - val_f1_score: 0.7423
Epoch 14/50
9819/9819 [==============================] - 103s 10ms/step - loss: 0.2714 - categorical_accuracy: 0.9626 - precision: 0.9669 - recall: 0.9608 - f1_score: 0.7835 - val_loss: 0.2162 - val_categorical_accuracy: 0.9368 - val_precision: 0.9404 - val_recall: 0.9339 - val_f1_score: 0.7481
Epoch 15/50
9819/9819 [==============================] - 103s 10ms/step - loss: 0.2688 - categorical_accuracy: 0.9643 - precision: 0.9686 - recall: 0.9630 - f1_score: 0.7846 - val_loss: 0.2437 - val_categorical_accuracy: 0.9325 - val_precision: 0.9354 - val_recall: 0.9307 - val_f1_score: 0.7500
Epoch 16/50
9819/9819 [==============================] - 103s 11ms/step - loss: 0.2657 - categorical_accuracy: 0.9649 - precision: 0.9690 - recall: 0.9635 - f1_score: 0.7851 - val_loss: 0.2073 - val_categorical_accuracy: 0.9371 - val_precision: 0.9400 - val_recall: 0.9351 - val_f1_score: 0.7446
Epoch 17/50
9819/9819 [==============================] - 105s 11ms/step - loss: 0.2643 - categorical_accuracy: 0.9661 - precision: 0.9707 - recall: 0.9646 - f1_score: 0.7874 - val_loss: 0.2357 - val_categorical_accuracy: 0.9350 - val_precision: 0.9376 - val_recall: 0.9331 - val_f1_score: 0.7498
Epoch 18/50
9819/9819 [==============================] - 103s 11ms/step - loss: 0.2574 - categorical_accuracy: 0.9670 - precision: 0.9709 - recall: 0.9658 - f1_score: 0.7884 - val_loss: 0.2125 - val_categorical_accuracy: 0.9379 - val_precision: 0.9405 - val_recall: 0.9363 - val_f1_score: 0.7474
Epoch 19/50
9819/9819 [==============================] - 103s 10ms/step - loss: 0.2578 - categorical_accuracy: 0.9683 - precision: 0.9720 - recall: 0.9674 - f1_score: 0.7889 - val_loss: 0.2048 - val_categorical_accuracy: 0.9391 - val_precision: 0.9411 - val_recall: 0.9377 - val_f1_score: 0.7556
Epoch 20/50
9819/9819 [==============================] - 103s 10ms/step - loss: 0.2566 - categorical_accuracy: 0.9688 - precision: 0.9725 - recall: 0.9678 - f1_score: 0.7895 - val_loss: 0.1768 - val_categorical_accuracy: 0.9472 - val_precision: 0.9486 - val_recall: 0.9457 - val_f1_score: 0.7536
Epoch 21/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2529 - categorical_accuracy: 0.9698 - precision: 0.9734 - recall: 0.9688 - f1_score: 0.7906 - val_loss: 0.2140 - val_categorical_accuracy: 0.9404 - val_precision: 0.9422 - val_recall: 0.9391 - val_f1_score: 0.7526
Epoch 22/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.2542 - categorical_accuracy: 0.9711 - precision: 0.9748 - recall: 0.9703 - f1_score: 0.7918 - val_loss: 0.2228 - val_categorical_accuracy: 0.9395 - val_precision: 0.9423 - val_recall: 0.9377 - val_f1_score: 0.7505
Epoch 23/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2515 - categorical_accuracy: 0.9720 - precision: 0.9756 - recall: 0.9713 - f1_score: 0.7920 - val_loss: 0.2329 - val_categorical_accuracy: 0.9356 - val_precision: 0.9380 - val_recall: 0.9337 - val_f1_score: 0.7490
Epoch 24/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.2490 - categorical_accuracy: 0.9724 - precision: 0.9759 - recall: 0.9717 - f1_score: 0.7921 - val_loss: 0.2098 - val_categorical_accuracy: 0.9448 - val_precision: 0.9471 - val_recall: 0.9437 - val_f1_score: 0.7560
Epoch 25/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2532 - categorical_accuracy: 0.9732 - precision: 0.9766 - recall: 0.9724 - f1_score: 0.7921 - val_loss: 0.2465 - val_categorical_accuracy: 0.9370 - val_precision: 0.9384 - val_recall: 0.9355 - val_f1_score: 0.7556
Epoch 26/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2451 - categorical_accuracy: 0.9730 - precision: 0.9764 - recall: 0.9724 - f1_score: 0.7925 - val_loss: 0.2174 - val_categorical_accuracy: 0.9425 - val_precision: 0.9448 - val_recall: 0.9412 - val_f1_score: 0.7562
Epoch 27/50
9819/9819 [==============================] - 105s 11ms/step - loss: 0.2487 - categorical_accuracy: 0.9728 - precision: 0.9763 - recall: 0.9723 - f1_score: 0.7936 - val_loss: 0.2401 - val_categorical_accuracy: 0.9384 - val_precision: 0.9404 - val_recall: 0.9373 - val_f1_score: 0.7556
Epoch 28/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2447 - categorical_accuracy: 0.9742 - precision: 0.9776 - recall: 0.9735 - f1_score: 0.7952 - val_loss: 0.2352 - val_categorical_accuracy: 0.9388 - val_precision: 0.9408 - val_recall: 0.9371 - val_f1_score: 0.7515
Epoch 29/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2448 - categorical_accuracy: 0.9740 - precision: 0.9774 - recall: 0.9734 - f1_score: 0.7930 - val_loss: 0.2131 - val_categorical_accuracy: 0.9457 - val_precision: 0.9470 - val_recall: 0.9448 - val_f1_score: 0.7581
Epoch 30/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2441 - categorical_accuracy: 0.9746 - precision: 0.9779 - recall: 0.9741 - f1_score: 0.7940 - val_loss: 0.2012 - val_categorical_accuracy: 0.9495 - val_precision: 0.9509 - val_recall: 0.9484 - val_f1_score: 0.7585
Epoch 31/50
9819/9819 [==============================] - 102s 10ms/step - loss: 0.2453 - categorical_accuracy: 0.9749 - precision: 0.9782 - recall: 0.9744 - f1_score: 0.7958 - val_loss: 0.2544 - val_categorical_accuracy: 0.9400 - val_precision: 0.9417 - val_recall: 0.9383 - val_f1_score: 0.7575
Epoch 32/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.2367 - categorical_accuracy: 0.9762 - precision: 0.9795 - recall: 0.9757 - f1_score: 0.7985 - val_loss: 0.2474 - val_categorical_accuracy: 0.9416 - val_precision: 0.9439 - val_recall: 0.9402 - val_f1_score: 0.7578
Epoch 33/50
9819/9819 [==============================] - 105s 11ms/step - loss: 0.2434 - categorical_accuracy: 0.9760 - precision: 0.9794 - recall: 0.9755 - f1_score: 0.7957 - val_loss: 0.2404 - val_categorical_accuracy: 0.9419 - val_precision: 0.9435 - val_recall: 0.9409 - val_f1_score: 0.7555
Epoch 34/50
9819/9819 [==============================] - 103s 11ms/step - loss: 0.2384 - categorical_accuracy: 0.9763 - precision: 0.9797 - recall: 0.9756 - f1_score: 0.7987 - val_loss: 0.2101 - val_categorical_accuracy: 0.9476 - val_precision: 0.9494 - val_recall: 0.9468 - val_f1_score: 0.7571
Epoch 35/50
9819/9819 [==============================] - 104s 11ms/step - loss: 0.2426 - categorical_accuracy: 0.9766 - precision: 0.9799 - recall: 0.9762 - f1_score: 0.7966 - val_loss: 0.2265 - val_categorical_accuracy: 0.9470 - val_precision: 0.9483 - val_recall: 0.9460 - val_f1_score: 0.7610
Epoch 36/50
9819/9819 [==============================] - 105s 11ms/step - loss: 0.2476 - categorical_accuracy: 0.9766 - precision: 0.9800 - recall: 0.9762 - f1_score: 0.7981 - val_loss: 0.2243 - val_categorical_accuracy: 0.9423 - val_precision: 0.9437 - val_recall: 0.9412 - val_f1_score: 0.7525
Epoch 37/50
9819/9819 [==============================] - 107s 11ms/step - loss: 0.2374 - categorical_accuracy: 0.9779 - precision: 0.9812 - recall: 0.9775 - f1_score: 0.7994 - val_loss: 0.2419 - val_categorical_accuracy: 0.9421 - val_precision: 0.9434 - val_recall: 0.9414 - val_f1_score: 0.7553
Epoch 38/50
9819/9819 [==============================] - 109s 11ms/step - loss: 0.2369 - categorical_accuracy: 0.9774 - precision: 0.9807 - recall: 0.9771 - f1_score: 0.7991 - val_loss: 0.2362 - val_categorical_accuracy: 0.9450 - val_precision: 0.9467 - val_recall: 0.9438 - val_f1_score: 0.7547
Epoch 39/50
9819/9819 [==============================] - 105s 11ms/step - loss: 0.2367 - categorical_accuracy: 0.9780 - precision: 0.9813 - recall: 0.9776 - f1_score: 0.7973 - val_loss: 0.2373 - val_categorical_accuracy: 0.9439 - val_precision: 0.9453 - val_recall: 0.9428 - val_f1_score: 0.7562
Epoch 40/50
9819/9819 [==============================] - 106s 11ms/step - loss: 0.2344 - categorical_accuracy: 0.9783 - precision: 0.9816 - recall: 0.9779 - f1_score: 0.8005 - val_loss: 0.2433 - val_categorical_accuracy: 0.9416 - val_precision: 0.9434 - val_recall: 0.9407 - val_f1_score: 0.7543

Val Classification Report
              precision    recall  f1-score   support

           0       0.97      0.92      0.94     12978
           1       0.84      0.98      0.90      3780
           2       1.00      0.98      0.99       360
           3       0.99      0.93      0.96      4650
           4       0.83      0.88      0.86      2539
           5       0.98      0.98      0.98     14123
           6       1.00      0.98      0.99      5400
           7       0.00      0.00      0.00        14
           8       0.90      0.98      0.94       748
         101       0.92      0.90      0.91      7435
         102       0.65      0.98      0.78        81
         105       0.87      0.98      0.92      2004
         106       0.96      0.99      0.98      2160
         107       0.77      0.85      0.80       635
         108       0.98      0.95      0.97      1772

    accuracy                           0.95     58679
   macro avg       0.84      0.89      0.86     58679
weighted avg       0.95      0.95      0.95     58679

Test Classification Report
              precision    recall  f1-score   support

           0       0.98      0.93      0.95     33789
           1       0.88      0.93      0.90      9399
           2       0.94      0.95      0.95      1097
           3       0.99      0.98      0.98     15400
           4       0.81      0.99      0.89      6246
           5       0.99      0.97      0.98     35960
           6       0.99      0.98      0.98     12914
           7       0.00      0.00      0.00        87
           8       0.81      0.99      0.89      1746
         101       0.93      0.91      0.92     18727
         102       0.74      0.77      0.75       452
         105       0.85      0.97      0.91      5151
         106       0.95      0.97      0.96      5173
         107       0.95      0.99      0.97     11580
         108       0.98      0.92      0.95      4974

    accuracy                           0.95    162695
   macro avg       0.85      0.88      0.87    162695
weighted avg       0.96      0.95      0.95    162695

All Data Classification Report
              precision    recall  f1-score   support

           0       0.98      0.96      0.97    164820
           1       0.93      0.97      0.95     48436
           2       0.97      0.98      0.98      6042
           3       1.00      0.99      0.99     77554
           4       0.93      0.99      0.96     31211
           5       1.00      0.98      0.99    176186
           6       0.99      0.98      0.99     64648
           7       0.00      0.00      0.00      1815
           8       0.91      0.99      0.95     10059
         101       0.95      0.95      0.95     87937
         102       0.78      0.92      0.84      2339
         105       0.91      0.99      0.95     31150
         106       0.96      0.98      0.97     25955
         107       0.97      0.99      0.98     38338
         108       0.99      0.96      0.97     23961

    accuracy                           0.97    790451
   macro avg       0.89      0.91      0.90    790451
weighted avg       0.97      0.97      0.97    790451

model weights saved in rnn_model_batch_size_64-seq_length_30-2022-09-08T19_48_42.h5
Wall time: 1h 13min 11s

CNN model

class CNNModel(tf.keras.Model):

  def __init__(self):
    super().__init__()
    self.norm = layers.LayerNormalization()
    self.conv1 = layers.Conv1D(filters=128, kernel_size=2, padding="same", activation='relu', name='conv1')
    self.MP1 = layers.MaxPooling1D(pool_size=2, name='MP1')
    self.drop1 = layers.Dropout(0.2, seed=200560, name='drop1')
    
    self.conv2 = layers.Conv1D(filters=128, kernel_size=2, padding="same", activation='relu', name='conv2')
    self.MP2 = layers.MaxPooling1D(pool_size=2, name='MP2')
    self.drop2 = layers.Dropout(0.2, seed=200560, name='drop2')
    
    self.conv3 = layers.Conv1D(filters=128, kernel_size=2, padding="same", activation='relu', name='conv3')
    self.MP3 = layers.MaxPooling1D(pool_size=2, name='MP3')
    self.drop3 = layers.Dropout(0.2, seed=200560, name='drop3')
    
    self.flatten = layers.Flatten()
    
    self.dense = layers.Dense(17, activation="relu", name='dense')
    self.softmax = layers.Softmax()
    return 

  def call(self, inputs, training=False):
    x = self.norm(inputs)
    x = self.conv1(x)
    x = self.MP1(x)
    x = self.drop1(x, training=training)
    
    x = self.conv2(x)
    x = self.MP2(x)
    x = self.drop2(x, training=training)
    
    x = self.conv3(x)
    x = self.MP3(x)
    x = self.drop3(x, training=training)
    
    x = self.flatten(x)
    
    x = self.dense(x)
    
    return self.softmax(x)

%%time

train_df, test_df, val_df = dset.split(real=True, simul=True, drawn=True, test_size=0.2, val_size=0.1)

tf.keras.backend.clear_session()
tf.compat.v1.reset_default_graph()

cnn_model = kmodel('CNNModel', CNNModel, flist0, 'class', categories, 64, 30, 
                      train_df, val_df, test_df, 'D:/datatmp', reset_ts=False, class_bal=True)

cnn_model.compile_and_fit(max_epochs=50, patience=10, lr=0.0001)

cnn_model.save()

Instances Train: 1425  Test: 397  Validation: 159
Model: "cnn_model"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 layer_normalization (LayerN  (64, 30, 16)             32        
 ormalization)                                                   
                                                                 
 conv1 (Conv1D)              (64, 30, 128)             4224      
                                                                 
 MP1 (MaxPooling1D)          (64, 15, 128)             0         
                                                                 
 drop1 (Dropout)             (64, 15, 128)             0         
                                                                 
 conv2 (Conv1D)              (64, 15, 128)             32896     
                                                                 
 MP2 (MaxPooling1D)          (64, 7, 128)              0         
                                                                 
 drop2 (Dropout)             (64, 7, 128)              0         
                                                                 
 conv3 (Conv1D)              (64, 7, 128)              32896     
                                                                 
 MP3 (MaxPooling1D)          (64, 3, 128)              0         
                                                                 
 drop3 (Dropout)             (64, 3, 128)              0         
                                                                 
 flatten (Flatten)           (64, 384)                 0         
                                                                 
 dense (Dense)               (64, 17)                  6545      
                                                                 
 softmax (Softmax)           (64, 17)                  0         
                                                                 
=================================================================
Total params: 76,593
Trainable params: 76,593
Non-trainable params: 0
_________________________________________________________________
None
Epoch 1/50
9819/9819 [==============================] - 184s 19ms/step - loss: 1.3076 - categorical_accuracy: 0.6411 - precision: 0.8336 - recall: 0.4975 - f1_score: 0.4811 - val_loss: 0.7030 - val_categorical_accuracy: 0.7467 - val_precision: 0.8424 - val_recall: 0.6463 - val_f1_score: 0.5976
Epoch 2/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.6757 - categorical_accuracy: 0.8051 - precision: 0.8613 - recall: 0.7486 - f1_score: 0.6418 - val_loss: 0.4856 - val_categorical_accuracy: 0.8363 - val_precision: 0.8810 - val_recall: 0.7658 - val_f1_score: 0.6334
Epoch 3/50
9819/9819 [==============================] - 106s 11ms/step - loss: 0.5388 - categorical_accuracy: 0.8489 - precision: 0.8847 - recall: 0.8124 - f1_score: 0.6823 - val_loss: 0.3886 - val_categorical_accuracy: 0.8725 - val_precision: 0.9047 - val_recall: 0.8279 - val_f1_score: 0.6620
Epoch 4/50
9819/9819 [==============================] - 115s 12ms/step - loss: 0.4848 - categorical_accuracy: 0.8702 - precision: 0.8972 - recall: 0.8431 - f1_score: 0.7004 - val_loss: 0.3660 - val_categorical_accuracy: 0.8739 - val_precision: 0.8945 - val_recall: 0.8436 - val_f1_score: 0.6792
Epoch 5/50
9819/9819 [==============================] - 115s 12ms/step - loss: 0.4437 - categorical_accuracy: 0.8853 - precision: 0.9069 - recall: 0.8639 - f1_score: 0.7148 - val_loss: 0.3067 - val_categorical_accuracy: 0.8958 - val_precision: 0.9100 - val_recall: 0.8783 - val_f1_score: 0.6932
Epoch 6/50
9819/9819 [==============================] - 121s 12ms/step - loss: 0.4240 - categorical_accuracy: 0.8941 - precision: 0.9131 - recall: 0.8767 - f1_score: 0.7226 - val_loss: 0.2704 - val_categorical_accuracy: 0.9032 - val_precision: 0.9169 - val_recall: 0.8903 - val_f1_score: 0.7014
Epoch 7/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.4021 - categorical_accuracy: 0.9030 - precision: 0.9198 - recall: 0.8877 - f1_score: 0.7316 - val_loss: 0.2809 - val_categorical_accuracy: 0.9011 - val_precision: 0.9157 - val_recall: 0.8883 - val_f1_score: 0.7177
Epoch 8/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.3864 - categorical_accuracy: 0.9076 - precision: 0.9222 - recall: 0.8944 - f1_score: 0.7361 - val_loss: 0.2520 - val_categorical_accuracy: 0.9126 - val_precision: 0.9260 - val_recall: 0.8979 - val_f1_score: 0.7198
Epoch 9/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.3761 - categorical_accuracy: 0.9148 - precision: 0.9276 - recall: 0.9033 - f1_score: 0.7406 - val_loss: 0.2730 - val_categorical_accuracy: 0.9029 - val_precision: 0.9168 - val_recall: 0.8844 - val_f1_score: 0.7085
Epoch 10/50
9819/9819 [==============================] - 109s 11ms/step - loss: 0.3649 - categorical_accuracy: 0.9171 - precision: 0.9299 - recall: 0.9061 - f1_score: 0.7428 - val_loss: 0.2332 - val_categorical_accuracy: 0.9186 - val_precision: 0.9290 - val_recall: 0.9064 - val_f1_score: 0.7260
Epoch 11/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.3524 - categorical_accuracy: 0.9211 - precision: 0.9320 - recall: 0.9111 - f1_score: 0.7465 - val_loss: 0.2706 - val_categorical_accuracy: 0.9034 - val_precision: 0.9155 - val_recall: 0.8884 - val_f1_score: 0.7026
Epoch 12/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.3507 - categorical_accuracy: 0.9258 - precision: 0.9363 - recall: 0.9168 - f1_score: 0.7507 - val_loss: 0.2281 - val_categorical_accuracy: 0.9121 - val_precision: 0.9206 - val_recall: 0.8995 - val_f1_score: 0.7176
Epoch 13/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.3333 - categorical_accuracy: 0.9294 - precision: 0.9386 - recall: 0.9217 - f1_score: 0.7562 - val_loss: 0.2184 - val_categorical_accuracy: 0.9199 - val_precision: 0.9277 - val_recall: 0.9080 - val_f1_score: 0.7164
Epoch 14/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.3373 - categorical_accuracy: 0.9309 - precision: 0.9401 - recall: 0.9229 - f1_score: 0.7547 - val_loss: 0.2472 - val_categorical_accuracy: 0.9107 - val_precision: 0.9199 - val_recall: 0.8977 - val_f1_score: 0.7201
Epoch 15/50
9819/9819 [==============================] - 112s 11ms/step - loss: 0.3250 - categorical_accuracy: 0.9327 - precision: 0.9411 - recall: 0.9257 - f1_score: 0.7571 - val_loss: 0.1917 - val_categorical_accuracy: 0.9290 - val_precision: 0.9353 - val_recall: 0.9210 - val_f1_score: 0.7327
Epoch 16/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.3268 - categorical_accuracy: 0.9358 - precision: 0.9434 - recall: 0.9294 - f1_score: 0.7594 - val_loss: 0.2532 - val_categorical_accuracy: 0.9085 - val_precision: 0.9158 - val_recall: 0.8902 - val_f1_score: 0.7147
Epoch 17/50
9819/9819 [==============================] - 113s 11ms/step - loss: 0.3191 - categorical_accuracy: 0.9372 - precision: 0.9447 - recall: 0.9318 - f1_score: 0.7613 - val_loss: 0.2085 - val_categorical_accuracy: 0.9241 - val_precision: 0.9308 - val_recall: 0.9121 - val_f1_score: 0.7244
Epoch 18/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.3128 - categorical_accuracy: 0.9395 - precision: 0.9467 - recall: 0.9339 - f1_score: 0.7634 - val_loss: 0.2015 - val_categorical_accuracy: 0.9245 - val_precision: 0.9309 - val_recall: 0.9159 - val_f1_score: 0.7149
Epoch 19/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.3126 - categorical_accuracy: 0.9417 - precision: 0.9484 - recall: 0.9366 - f1_score: 0.7645 - val_loss: 0.1984 - val_categorical_accuracy: 0.9278 - val_precision: 0.9330 - val_recall: 0.9200 - val_f1_score: 0.7345
Epoch 20/50
9819/9819 [==============================] - 107s 11ms/step - loss: 0.3060 - categorical_accuracy: 0.9429 - precision: 0.9493 - recall: 0.9386 - f1_score: 0.7670 - val_loss: 0.2007 - val_categorical_accuracy: 0.9295 - val_precision: 0.9330 - val_recall: 0.9203 - val_f1_score: 0.7359
Epoch 21/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.3016 - categorical_accuracy: 0.9431 - precision: 0.9493 - recall: 0.9389 - f1_score: 0.7673 - val_loss: 0.2094 - val_categorical_accuracy: 0.9267 - val_precision: 0.9325 - val_recall: 0.9175 - val_f1_score: 0.7286
Epoch 22/50
9819/9819 [==============================] - 112s 11ms/step - loss: 0.3021 - categorical_accuracy: 0.9452 - precision: 0.9512 - recall: 0.9416 - f1_score: 0.7698 - val_loss: 0.1704 - val_categorical_accuracy: 0.9418 - val_precision: 0.9455 - val_recall: 0.9344 - val_f1_score: 0.7419
Epoch 23/50
9819/9819 [==============================] - 112s 11ms/step - loss: 0.2990 - categorical_accuracy: 0.9471 - precision: 0.9529 - recall: 0.9436 - f1_score: 0.7689 - val_loss: 0.1904 - val_categorical_accuracy: 0.9364 - val_precision: 0.9399 - val_recall: 0.9297 - val_f1_score: 0.7388
Epoch 24/50
9819/9819 [==============================] - 107s 11ms/step - loss: 0.2935 - categorical_accuracy: 0.9483 - precision: 0.9540 - recall: 0.9445 - f1_score: 0.7721 - val_loss: 0.2310 - val_categorical_accuracy: 0.9139 - val_precision: 0.9186 - val_recall: 0.9072 - val_f1_score: 0.7180
Epoch 25/50
9819/9819 [==============================] - 106s 11ms/step - loss: 0.2943 - categorical_accuracy: 0.9495 - precision: 0.9552 - recall: 0.9464 - f1_score: 0.7731 - val_loss: 0.1633 - val_categorical_accuracy: 0.9379 - val_precision: 0.9427 - val_recall: 0.9320 - val_f1_score: 0.7462
Epoch 26/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.2909 - categorical_accuracy: 0.9499 - precision: 0.9552 - recall: 0.9470 - f1_score: 0.7722 - val_loss: 0.1689 - val_categorical_accuracy: 0.9422 - val_precision: 0.9449 - val_recall: 0.9359 - val_f1_score: 0.7395
Epoch 27/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.2929 - categorical_accuracy: 0.9512 - precision: 0.9565 - recall: 0.9488 - f1_score: 0.7736 - val_loss: 0.1887 - val_categorical_accuracy: 0.9344 - val_precision: 0.9412 - val_recall: 0.9282 - val_f1_score: 0.7148
Epoch 28/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.2850 - categorical_accuracy: 0.9526 - precision: 0.9579 - recall: 0.9499 - f1_score: 0.7760 - val_loss: 0.1928 - val_categorical_accuracy: 0.9270 - val_precision: 0.9306 - val_recall: 0.9224 - val_f1_score: 0.7361
Epoch 29/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.2926 - categorical_accuracy: 0.9530 - precision: 0.9579 - recall: 0.9503 - f1_score: 0.7765 - val_loss: 0.1627 - val_categorical_accuracy: 0.9437 - val_precision: 0.9467 - val_recall: 0.9397 - val_f1_score: 0.7409
Epoch 30/50
9819/9819 [==============================] - 112s 11ms/step - loss: 0.2836 - categorical_accuracy: 0.9546 - precision: 0.9594 - recall: 0.9522 - f1_score: 0.7766 - val_loss: 0.1625 - val_categorical_accuracy: 0.9424 - val_precision: 0.9464 - val_recall: 0.9386 - val_f1_score: 0.7432
Epoch 31/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.2820 - categorical_accuracy: 0.9544 - precision: 0.9593 - recall: 0.9523 - f1_score: 0.7756 - val_loss: 0.1828 - val_categorical_accuracy: 0.9362 - val_precision: 0.9406 - val_recall: 0.9316 - val_f1_score: 0.7251
Epoch 32/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2778 - categorical_accuracy: 0.9556 - precision: 0.9603 - recall: 0.9537 - f1_score: 0.7781 - val_loss: 0.1589 - val_categorical_accuracy: 0.9446 - val_precision: 0.9476 - val_recall: 0.9411 - val_f1_score: 0.7485
Epoch 33/50
9819/9819 [==============================] - 107s 11ms/step - loss: 0.2782 - categorical_accuracy: 0.9558 - precision: 0.9607 - recall: 0.9539 - f1_score: 0.7775 - val_loss: 0.1799 - val_categorical_accuracy: 0.9424 - val_precision: 0.9459 - val_recall: 0.9389 - val_f1_score: 0.7451
Epoch 34/50
9819/9819 [==============================] - 107s 11ms/step - loss: 0.2740 - categorical_accuracy: 0.9566 - precision: 0.9613 - recall: 0.9548 - f1_score: 0.7788 - val_loss: 0.1643 - val_categorical_accuracy: 0.9399 - val_precision: 0.9448 - val_recall: 0.9350 - val_f1_score: 0.7380
Epoch 35/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2722 - categorical_accuracy: 0.9579 - precision: 0.9627 - recall: 0.9561 - f1_score: 0.7809 - val_loss: 0.1827 - val_categorical_accuracy: 0.9363 - val_precision: 0.9430 - val_recall: 0.9293 - val_f1_score: 0.7453
Epoch 36/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2751 - categorical_accuracy: 0.9577 - precision: 0.9625 - recall: 0.9560 - f1_score: 0.7796 - val_loss: 0.1629 - val_categorical_accuracy: 0.9440 - val_precision: 0.9472 - val_recall: 0.9410 - val_f1_score: 0.7462
Epoch 37/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2712 - categorical_accuracy: 0.9585 - precision: 0.9630 - recall: 0.9570 - f1_score: 0.7801 - val_loss: 0.1624 - val_categorical_accuracy: 0.9422 - val_precision: 0.9468 - val_recall: 0.9381 - val_f1_score: 0.7306
Epoch 38/50
9819/9819 [==============================] - 109s 11ms/step - loss: 0.2723 - categorical_accuracy: 0.9595 - precision: 0.9639 - recall: 0.9577 - f1_score: 0.7811 - val_loss: 0.1837 - val_categorical_accuracy: 0.9378 - val_precision: 0.9405 - val_recall: 0.9348 - val_f1_score: 0.7446
Epoch 39/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2708 - categorical_accuracy: 0.9602 - precision: 0.9646 - recall: 0.9586 - f1_score: 0.7820 - val_loss: 0.1598 - val_categorical_accuracy: 0.9396 - val_precision: 0.9446 - val_recall: 0.9365 - val_f1_score: 0.7414
Epoch 40/50
9819/9819 [==============================] - 112s 11ms/step - loss: 0.2672 - categorical_accuracy: 0.9596 - precision: 0.9643 - recall: 0.9582 - f1_score: 0.7829 - val_loss: 0.1601 - val_categorical_accuracy: 0.9430 - val_precision: 0.9478 - val_recall: 0.9382 - val_f1_score: 0.7433
Epoch 41/50
9819/9819 [==============================] - 112s 11ms/step - loss: 0.2652 - categorical_accuracy: 0.9612 - precision: 0.9658 - recall: 0.9598 - f1_score: 0.7842 - val_loss: 0.1664 - val_categorical_accuracy: 0.9492 - val_precision: 0.9535 - val_recall: 0.9452 - val_f1_score: 0.7466
Epoch 42/50
9819/9819 [==============================] - 113s 11ms/step - loss: 0.2661 - categorical_accuracy: 0.9610 - precision: 0.9656 - recall: 0.9597 - f1_score: 0.7823 - val_loss: 0.1480 - val_categorical_accuracy: 0.9454 - val_precision: 0.9482 - val_recall: 0.9434 - val_f1_score: 0.7514
Epoch 43/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2682 - categorical_accuracy: 0.9612 - precision: 0.9656 - recall: 0.9601 - f1_score: 0.7828 - val_loss: 0.1591 - val_categorical_accuracy: 0.9474 - val_precision: 0.9523 - val_recall: 0.9409 - val_f1_score: 0.7458
Epoch 44/50
9819/9819 [==============================] - 109s 11ms/step - loss: 0.2617 - categorical_accuracy: 0.9625 - precision: 0.9668 - recall: 0.9613 - f1_score: 0.7847 - val_loss: 0.1602 - val_categorical_accuracy: 0.9473 - val_precision: 0.9515 - val_recall: 0.9377 - val_f1_score: 0.7435
Epoch 45/50
9819/9819 [==============================] - 109s 11ms/step - loss: 0.2603 - categorical_accuracy: 0.9629 - precision: 0.9673 - recall: 0.9618 - f1_score: 0.7860 - val_loss: 0.1615 - val_categorical_accuracy: 0.9439 - val_precision: 0.9472 - val_recall: 0.9406 - val_f1_score: 0.7461
Epoch 46/50
9819/9819 [==============================] - 109s 11ms/step - loss: 0.2647 - categorical_accuracy: 0.9623 - precision: 0.9666 - recall: 0.9612 - f1_score: 0.7838 - val_loss: 0.1443 - val_categorical_accuracy: 0.9472 - val_precision: 0.9508 - val_recall: 0.9440 - val_f1_score: 0.7413
Epoch 47/50
9819/9819 [==============================] - 107s 11ms/step - loss: 0.2617 - categorical_accuracy: 0.9638 - precision: 0.9681 - recall: 0.9624 - f1_score: 0.7855 - val_loss: 0.1685 - val_categorical_accuracy: 0.9421 - val_precision: 0.9466 - val_recall: 0.9383 - val_f1_score: 0.7410
Epoch 48/50
9819/9819 [==============================] - 108s 11ms/step - loss: 0.2644 - categorical_accuracy: 0.9637 - precision: 0.9680 - recall: 0.9625 - f1_score: 0.7846 - val_loss: 0.1739 - val_categorical_accuracy: 0.9424 - val_precision: 0.9459 - val_recall: 0.9390 - val_f1_score: 0.7418
Epoch 49/50
9819/9819 [==============================] - 110s 11ms/step - loss: 0.2598 - categorical_accuracy: 0.9640 - precision: 0.9682 - recall: 0.9628 - f1_score: 0.7866 - val_loss: 0.1655 - val_categorical_accuracy: 0.9473 - val_precision: 0.9504 - val_recall: 0.9450 - val_f1_score: 0.7470
Epoch 50/50
9819/9819 [==============================] - 111s 11ms/step - loss: 0.2574 - categorical_accuracy: 0.9652 - precision: 0.9693 - recall: 0.9640 - f1_score: 0.7878 - val_loss: 0.1595 - val_categorical_accuracy: 0.9404 - val_precision: 0.9445 - val_recall: 0.9380 - val_f1_score: 0.7426

Val Classification Report
              precision    recall  f1-score   support

           0       0.97      0.90      0.93     12978
           1       0.84      0.97      0.90      3780
           2       1.00      0.97      0.98       360
           3       0.99      0.96      0.97      4650
           4       0.90      0.88      0.89      2539
           5       1.00      0.98      0.99     14123
           6       1.00      0.98      0.99      5400
           7       0.00      0.00      0.00        14
           8       0.95      0.97      0.96       748
         101       0.89      0.90      0.90      7435
         102       0.52      0.83      0.64        81
         105       0.86      0.98      0.92      2004
         106       0.94      1.00      0.97      2160
         107       0.67      0.89      0.76       635
         108       0.98      0.98      0.98      1772

    accuracy                           0.95     58679
   macro avg       0.83      0.88      0.85     58679
weighted avg       0.95      0.95      0.95     58679

Test Classification Report
              precision    recall  f1-score   support

           0       0.98      0.88      0.93     33789
           1       0.90      0.89      0.90      9399
           2       0.93      0.99      0.96      1097
           3       0.99      0.98      0.99     15400
           4       0.85      0.98      0.91      6246
           5       1.00      0.98      0.99     35960
           6       0.99      0.97      0.98     12914
           7       0.00      0.00      0.00        87
           8       0.85      0.98      0.91      1746
         101       0.87      0.93      0.90     18727
         102       0.66      0.63      0.65       452
         105       0.82      0.97      0.89      5151
         106       0.92      0.98      0.95      5173
         107       0.89      1.00      0.94     11580
         108       0.98      0.94      0.96      4974

    accuracy                           0.95    162695
   macro avg       0.84      0.87      0.86    162695
weighted avg       0.95      0.95      0.95    162695

All Data Classification Report
              precision    recall  f1-score   support

           0       0.98      0.92      0.95    164820
           1       0.94      0.94      0.94     48436
           2       0.98      0.99      0.98      6042
           3       1.00      0.99      0.99     77554
           4       0.93      0.98      0.96     31211
           5       1.00      0.98      0.99    176186
           6       1.00      0.97      0.98     64648
           7       0.00      0.00      0.00      1815
           8       0.94      0.98      0.96     10059
         101       0.90      0.95      0.93     87937
         102       0.69      0.90      0.78      2339
         105       0.89      0.99      0.94     31150
         106       0.93      0.99      0.96     25955
         107       0.91      1.00      0.95     38338
         108       0.99      0.97      0.98     23961

    accuracy                           0.96    790451
   macro avg       0.87      0.90      0.89    790451
weighted avg       0.96      0.96      0.96    790451

model weights saved in cnn_model_batch_size_64-seq_length_30-2022-09-08T06_49_06.h5
Wall time: 1h 35min 59s

Mixed model

class MixedModel(tf.keras.Model):

  def __init__(self):
    super().__init__()
    
    # CNN
    self.norm = layers.LayerNormalization(name='CNN_norm')
    self.conv1 = layers.Conv1D(filters=128, kernel_size=2, padding="same", activation='relu', name='conv1')
    self.MP1 = layers.MaxPooling1D(pool_size=2, name='MP1')
    self.drop1 = layers.Dropout(0.2, seed=200560, name='drop1')
    
    self.conv2 = layers.Conv1D(filters=128, kernel_size=2, padding="same", activation='relu', name='conv2')
    self.MP2 = layers.MaxPooling1D(pool_size=2, name='MP2')
    self.drop2 = layers.Dropout(0.2, seed=200560, name='drop2')
    
    self.conv3 = layers.Conv1D(filters=128, kernel_size=2, padding="same", activation='relu', name='conv3')
    self.MP3 = layers.MaxPooling1D(pool_size=2, name='MP3')
    self.drop3 = layers.Dropout(0.2, seed=200560, name='drop3')
    
    self.flatten = layers.Flatten(name='CNN_flatten')
    
    #RNN
    self.lstm = layers.LSTM(128, return_sequences=False, dropout=0.15, name='lstm')

    #Dense
    self.dense = layers.Dense(17, activation="relu", name='dense')
    self.softmax = layers.Softmax(name='softmax')

  def call(self, inputs, training=False):
    x = self.norm(inputs)
    x = self.conv1(x)
    x = self.MP1(x)
    x = self.drop1(x, training=training)
    
    x = self.conv2(x)
    x = self.MP2(x)
    x = self.drop2(x, training=training)
    
    x = self.conv3(x)
    x = self.MP3(x)
    x = self.drop3(x, training=training)
    
    x = self.flatten(x)
    
    y = self.lstm(inputs, training=training)

    z = tf.concat([x, y], 1)
    z = self.dense(z)
    return self.softmax(z)

%%time

train_df, test_df, val_df = dset.split(real=True, simul=True, drawn=True, test_size=0.2, val_size=0.1)

tf.keras.backend.clear_session()
tf.compat.v1.reset_default_graph()

mixed_model = kmodel('MixedModel', MixedModel, flist0, 'class', categories, 64, 30, 
                      train_df, val_df, test_df, 'D:/datatmp', reset_ts=True, class_bal=True)

mixed_model.compile_and_fit(max_epochs=100, patience=10, lr=0.0001)

Instances Train: 1425  Test: 397  Validation: 159
Model: "mixed_model"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 CNN_norm (LayerNormalizatio  (64, 30, 16)             32        
 n)                                                              
                                                                 
 conv1 (Conv1D)              (64, 30, 128)             4224      
                                                                 
 MP1 (MaxPooling1D)          (64, 15, 128)             0         
                                                                 
 drop1 (Dropout)             (64, 15, 128)             0         
                                                                 
 conv2 (Conv1D)              (64, 15, 128)             32896     
                                                                 
 MP2 (MaxPooling1D)          (64, 7, 128)              0         
                                                                 
 drop2 (Dropout)             (64, 7, 128)              0         
                                                                 
 conv3 (Conv1D)              (64, 7, 128)              32896     
                                                                 
 MP3 (MaxPooling1D)          (64, 3, 128)              0         
                                                                 
 drop3 (Dropout)             (64, 3, 128)              0         
                                                                 
 CNN_flatten (Flatten)       (64, 384)                 0         
                                                                 
 lstm (LSTM)                 (64, 128)                 74240     
                                                                 
 dense (Dense)               (64, 17)                  8721      
                                                                 
 softmax (Softmax)           (64, 17)                  0         
                                                                 
=================================================================
Total params: 153,009
Trainable params: 153,009
Non-trainable params: 0
_________________________________________________________________
None
Epoch 1/100
9819/9819 [==============================] - 330s 33ms/step - loss: 1.1338 - categorical_accuracy: 0.6731 - precision: 0.8550 - recall: 0.5317 - f1_score: 0.5241 - val_loss: 0.6545 - val_categorical_accuracy: 0.7712 - val_precision: 0.8519 - val_recall: 0.6655 - val_f1_score: 0.6060
Epoch 2/100
9819/9819 [==============================] - 148s 15ms/step - loss: 0.6045 - categorical_accuracy: 0.8280 - precision: 0.8779 - recall: 0.7757 - f1_score: 0.6662 - val_loss: 0.4053 - val_categorical_accuracy: 0.8731 - val_precision: 0.9049 - val_recall: 0.8276 - val_f1_score: 0.6622
Epoch 3/100
9819/9819 [==============================] - 150s 15ms/step - loss: 0.4973 - categorical_accuracy: 0.8690 - precision: 0.8998 - recall: 0.8394 - f1_score: 0.7031 - val_loss: 0.4333 - val_categorical_accuracy: 0.8570 - val_precision: 0.8961 - val_recall: 0.8002 - val_f1_score: 0.6580
Epoch 4/100
9819/9819 [==============================] - 150s 15ms/step - loss: 0.4357 - categorical_accuracy: 0.8899 - precision: 0.9122 - recall: 0.8675 - f1_score: 0.7204 - val_loss: 0.3056 - val_categorical_accuracy: 0.8898 - val_precision: 0.9080 - val_recall: 0.8652 - val_f1_score: 0.7064
Epoch 5/100
9819/9819 [==============================] - 148s 15ms/step - loss: 0.4137 - categorical_accuracy: 0.9050 - precision: 0.9223 - recall: 0.8873 - f1_score: 0.7305 - val_loss: 0.2479 - val_categorical_accuracy: 0.9188 - val_precision: 0.9294 - val_recall: 0.9087 - val_f1_score: 0.7121
Epoch 6/100
9819/9819 [==============================] - 151s 15ms/step - loss: 0.3772 - categorical_accuracy: 0.9157 - precision: 0.9297 - recall: 0.9020 - f1_score: 0.7424 - val_loss: 0.3250 - val_categorical_accuracy: 0.8777 - val_precision: 0.8923 - val_recall: 0.8579 - val_f1_score: 0.6920
Epoch 7/100
9819/9819 [==============================] - 154s 16ms/step - loss: 0.3638 - categorical_accuracy: 0.9214 - precision: 0.9332 - recall: 0.9108 - f1_score: 0.7456 - val_loss: 0.2710 - val_categorical_accuracy: 0.9088 - val_precision: 0.9197 - val_recall: 0.8888 - val_f1_score: 0.6993
Epoch 8/100
9819/9819 [==============================] - 148s 15ms/step - loss: 0.3461 - categorical_accuracy: 0.9270 - precision: 0.9377 - recall: 0.9176 - f1_score: 0.7534 - val_loss: 0.2629 - val_categorical_accuracy: 0.9123 - val_precision: 0.9227 - val_recall: 0.8975 - val_f1_score: 0.7001
Epoch 9/100
9819/9819 [==============================] - 146s 15ms/step - loss: 0.3487 - categorical_accuracy: 0.9307 - precision: 0.9402 - recall: 0.9220 - f1_score: 0.7537 - val_loss: 0.2489 - val_categorical_accuracy: 0.9097 - val_precision: 0.9206 - val_recall: 0.8995 - val_f1_score: 0.7028
Epoch 10/100
9819/9819 [==============================] - 152s 15ms/step - loss: 0.3250 - categorical_accuracy: 0.9351 - precision: 0.9434 - recall: 0.9281 - f1_score: 0.7605 - val_loss: 0.2429 - val_categorical_accuracy: 0.9164 - val_precision: 0.9249 - val_recall: 0.9053 - val_f1_score: 0.6989
Epoch 11/100
9819/9819 [==============================] - 155s 16ms/step - loss: 0.3206 - categorical_accuracy: 0.9394 - precision: 0.9473 - recall: 0.9328 - f1_score: 0.7632 - val_loss: 0.2105 - val_categorical_accuracy: 0.9247 - val_precision: 0.9326 - val_recall: 0.9192 - val_f1_score: 0.7342
Epoch 12/100
9819/9819 [==============================] - 153s 16ms/step - loss: 0.3125 - categorical_accuracy: 0.9421 - precision: 0.9491 - recall: 0.9366 - f1_score: 0.7653 - val_loss: 0.2176 - val_categorical_accuracy: 0.9235 - val_precision: 0.9290 - val_recall: 0.9148 - val_f1_score: 0.7221
Epoch 13/100
9819/9819 [==============================] - 153s 16ms/step - loss: 0.3111 - categorical_accuracy: 0.9455 - precision: 0.9521 - recall: 0.9401 - f1_score: 0.7670 - val_loss: 0.1829 - val_categorical_accuracy: 0.9329 - val_precision: 0.9393 - val_recall: 0.9280 - val_f1_score: 0.7286
Epoch 14/100
9819/9819 [==============================] - 147s 15ms/step - loss: 0.2964 - categorical_accuracy: 0.9484 - precision: 0.9544 - recall: 0.9443 - f1_score: 0.7729 - val_loss: 0.1681 - val_categorical_accuracy: 0.9419 - val_precision: 0.9462 - val_recall: 0.9388 - val_f1_score: 0.7332
Epoch 15/100
9819/9819 [==============================] - 146s 15ms/step - loss: 0.2933 - categorical_accuracy: 0.9499 - precision: 0.9557 - recall: 0.9462 - f1_score: 0.7737 - val_loss: 0.2068 - val_categorical_accuracy: 0.9276 - val_precision: 0.9328 - val_recall: 0.9212 - val_f1_score: 0.7374
Epoch 16/100
9819/9819 [==============================] - 150s 15ms/step - loss: 0.2948 - categorical_accuracy: 0.9523 - precision: 0.9580 - recall: 0.9485 - f1_score: 0.7738 - val_loss: 0.1975 - val_categorical_accuracy: 0.9318 - val_precision: 0.9368 - val_recall: 0.9259 - val_f1_score: 0.7257
Epoch 17/100
9819/9819 [==============================] - 151s 15ms/step - loss: 0.2893 - categorical_accuracy: 0.9543 - precision: 0.9595 - recall: 0.9511 - f1_score: 0.7764 - val_loss: 0.1754 - val_categorical_accuracy: 0.9410 - val_precision: 0.9464 - val_recall: 0.9363 - val_f1_score: 0.7517
Epoch 18/100
9819/9819 [==============================] - 150s 15ms/step - loss: 0.2827 - categorical_accuracy: 0.9558 - precision: 0.9607 - recall: 0.9528 - f1_score: 0.7777 - val_loss: 0.2170 - val_categorical_accuracy: 0.9245 - val_precision: 0.9301 - val_recall: 0.9198 - val_f1_score: 0.7105
Epoch 19/100
9819/9819 [==============================] - 151s 15ms/step - loss: 0.2763 - categorical_accuracy: 0.9565 - precision: 0.9615 - recall: 0.9540 - f1_score: 0.7798 - val_loss: 0.1697 - val_categorical_accuracy: 0.9406 - val_precision: 0.9441 - val_recall: 0.9373 - val_f1_score: 0.7375
Epoch 20/100
9819/9819 [==============================] - 152s 15ms/step - loss: 0.2749 - categorical_accuracy: 0.9586 - precision: 0.9633 - recall: 0.9566 - f1_score: 0.7817 - val_loss: 0.1589 - val_categorical_accuracy: 0.9437 - val_precision: 0.9468 - val_recall: 0.9412 - val_f1_score: 0.7384
Epoch 21/100
9819/9819 [==============================] - 160s 16ms/step - loss: 0.2747 - categorical_accuracy: 0.9597 - precision: 0.9644 - recall: 0.9577 - f1_score: 0.7815 - val_loss: 0.1510 - val_categorical_accuracy: 0.9483 - val_precision: 0.9520 - val_recall: 0.9445 - val_f1_score: 0.7385
Epoch 22/100
9819/9819 [==============================] - 150s 15ms/step - loss: 0.2730 - categorical_accuracy: 0.9603 - precision: 0.9649 - recall: 0.9581 - f1_score: 0.7817 - val_loss: 0.1769 - val_categorical_accuracy: 0.9418 - val_precision: 0.9461 - val_recall: 0.9361 - val_f1_score: 0.7378
Epoch 23/100
9819/9819 [==============================] - 151s 15ms/step - loss: 0.2701 - categorical_accuracy: 0.9624 - precision: 0.9666 - recall: 0.9605 - f1_score: 0.7836 - val_loss: 0.1881 - val_categorical_accuracy: 0.9336 - val_precision: 0.9377 - val_recall: 0.9308 - val_f1_score: 0.7313
Epoch 24/100
9819/9819 [==============================] - 148s 15ms/step - loss: 0.2684 - categorical_accuracy: 0.9622 - precision: 0.9665 - recall: 0.9605 - f1_score: 0.7843 - val_loss: 0.1649 - val_categorical_accuracy: 0.9433 - val_precision: 0.9461 - val_recall: 0.9403 - val_f1_score: 0.7476
Epoch 25/100
9819/9819 [==============================] - 147s 15ms/step - loss: 0.2679 - categorical_accuracy: 0.9637 - precision: 0.9680 - recall: 0.9620 - f1_score: 0.7858 - val_loss: 0.1847 - val_categorical_accuracy: 0.9391 - val_precision: 0.9426 - val_recall: 0.9358 - val_f1_score: 0.7393
Epoch 26/100
9819/9819 [==============================] - 148s 15ms/step - loss: 0.2692 - categorical_accuracy: 0.9648 - precision: 0.9689 - recall: 0.9632 - f1_score: 0.7843 - val_loss: 0.1695 - val_categorical_accuracy: 0.9445 - val_precision: 0.9473 - val_recall: 0.9419 - val_f1_score: 0.7435
Epoch 27/100
9819/9819 [==============================] - 150s 15ms/step - loss: 0.2582 - categorical_accuracy: 0.9653 - precision: 0.9694 - recall: 0.9640 - f1_score: 0.7873 - val_loss: 0.1615 - val_categorical_accuracy: 0.9443 - val_precision: 0.9477 - val_recall: 0.9416 - val_f1_score: 0.7399

Val Classification Report
              precision    recall  f1-score   support

           0       0.96      0.94      0.95     12978
           1       0.81      0.98      0.89      3780
           2       1.00      0.97      0.98       360
           3       0.94      0.97      0.95      4650
           4       0.86      0.90      0.88      2539
           5       0.99      0.95      0.97     14123
           6       0.99      0.96      0.98      5400
           7       0.00      0.00      0.00        14
           8       0.92      0.98      0.95       748
         101       0.98      0.85      0.91      7435
         102       0.59      0.99      0.74        81
         105       0.81      0.99      0.89      2004
         106       0.90      1.00      0.95      2160
         107       0.64      0.82      0.72       635
         108       0.99      0.96      0.98      1772

    accuracy                           0.94     58679
   macro avg       0.82      0.88      0.85     58679
weighted avg       0.95      0.94      0.94     58679

Test Classification Report
              precision    recall  f1-score   support

           0       0.96      0.92      0.94     33789
           1       0.87      0.96      0.92      9399
           2       0.95      1.00      0.97      1097
           3       0.98      0.99      0.98     15400
           4       0.83      0.97      0.89      6246
           5       1.00      0.96      0.98     35960
           6       0.99      0.96      0.97     12914
           7       0.00      0.00      0.00        87
           8       0.85      0.98      0.91      1746
         101       0.97      0.87      0.91     18727
         102       0.75      0.75      0.75       452
         105       0.71      0.98      0.83      5151
         106       0.90      0.98      0.94      5173
         107       0.92      1.00      0.96     11580
         108       0.98      0.94      0.96      4974

    accuracy                           0.95    162695
   macro avg       0.84      0.88      0.86    162695
weighted avg       0.95      0.95      0.95    162695

All Data Classification Report
              precision    recall  f1-score   support

           0       0.96      0.94      0.95    164820
           1       0.89      0.97      0.93     48436
           2       0.97      0.98      0.98      6042
           3       0.97      0.99      0.98     77554
           4       0.89      0.98      0.93     31211
           5       1.00      0.96      0.98    176186
           6       1.00      0.96      0.98     64648
           7       0.00      0.00      0.00      1815
           8       0.93      0.97      0.95     10059
         101       0.97      0.89      0.93     87937
         102       0.73      0.92      0.81      2339
         105       0.82      1.00      0.90     31150
         106       0.90      0.99      0.95     25955
         107       0.92      0.99      0.95     38338
         108       0.98      0.97      0.98     23961

    accuracy                           0.95    790451
   macro avg       0.86      0.90      0.88    790451
weighted avg       0.96      0.95      0.95    790451

Wall time: 1h 15min

Ancillary

dset_g = CustomDataGen(dset.drawn, flist0, 'class', categories, 64, 30, 'D:\datatmp')
fmodel = kmodel_from_file(RNNModel, (30, 16), 'RNNModel_batch_size_64-seq_length_30-2022-09-03T15_47_05.h5')

#dset_g.plot(0)
rnn_model.class_rep(dset_g)